Форум » ПЕТРОФИЗИЧЕСКИЙ СОФТ - PETROPHYSICAL SOFTWARE » Статистические задачи и алгоритмы » Ответить

Статистические задачи и алгоритмы

bne: Наверное пора сделать такой раздел Наиболее интересные ссылки по прикладной статистике можно размещать тут

Ответов - 116

bne: Neural Processing Letters Volume 27, Number 1, 57-66, DOI: 10.1007/s11063-007-9059-4 A Semi-supervised Learning Algorithm on Gaussian Mixture with Automatic Model Selection Zhiwu Lu and Yuxin Peng In Gaussian mixture modeling, it is crucial to select the number of Gaussians for a sample set, which becomes much more difficult when the overlap in the mixture is larger. Under regularization theory, we aim to solve this problem using a semi-supervised learning algorithm through incorporating pairwise constraints into entropy regularized likelihood (ERL) learning which can make automatic model selection for Gaussian mixture. The simulation experiments further demonstrate that the presented semi-supervised learning algorithm (i.e., the constrained ERL learning algorithm) can automatically detect the number of Gaussians with a good parameter estimation, even when two or more actual Gaussians in the mixture are overlapped at a high degree. Moreover, the constrained ERL learning algorithm leads to some promising results when applied to iris data classification and image database categorization. Keywords Entropy regularization - Gaussian mixture - Model selection - Pairwise constraints - Semi-supervised learning

bne: PCA Beyond The Concept of Manifolds: Principal Trees, Metro Maps, and Elastic Cubic Complexes Authors: A. N. Gorban, N. R. Sumner, A. Y. Zinovyev (Submitted on 31 Dec 2007) Abstract: Multidimensional data distributions can have complex topologies and variable local dimensions. To approximate complex data, we propose a new type of low-dimensional ``principal object'': a principal cubic complex. This complex is a generalization of linear and non-linear principal manifolds and includes them as a particular case. To construct such an object, we combine a method of topological grammars with the minimization of an elastic energy defined for its embedment into multidimensional data space. The whole complex is presented as a system of nodes and springs and as a product of one-dimensional continua (represented by graphs), and the grammars describe how these continua transform during the process of optimal complex construction. The simplest case of a topological grammar (``add a node'', ``bisect an edge'') is equivalent to the construction of ``principal trees'', an object useful in many practical applications. We demonstrate how it can be applied to the analysis of bacterial genomes and for visualization of cDNA microarray data using the ``metro map'' representation. The preprint is supplemented by animation: ``How the topological grammar constructs branching principal components (AnimatedBranchingPCA.gif)''. cтатья в PDF

bne: Energy Discriminant Analysis, Quantum Logic, and Fuzzy sets Authors: Grigorii Melnichenko (Submitted on 9 Nov 2007 (v1), last revised 28 Dec 2007 (this version, v2)) Abstract: In the paper, we show that quantum logic of linear subspaces can be used for recognition of random signals by a Bayesian energy discriminant classifier. The energy distribution on linear subspaces is described by the correlation matrix of the probability distribution. We show that the correlation matrix corresponds to von Neumann density matrix in quantum theory. We suggest the interpretation of quantum logic as a fuzzy logic of fuzzy sets. The use of quantum logic for recognition is based on the fact that the probability distribution of each class lies approximately in a lower-dimensional subspace of feature space. We offer the interpretation of discriminant functions as membership functions of fuzzy sets. Also we offer the quality functional for optimal choice of discriminant functions for recognition from some class of discriminant functions. Статья в PDF


lemirg: Я являюсь автором этой статьи. Любопытно, чем она првлекла этот форум. Я лично проверил применение этого метода для распознавания фонем из TIMIT базы. Вот только что получил разгромную рецензию на эту статью. Очень стало обидно. Рецензия CORRECTION OR COMMENTS FOR AUTHOR(S) Pattern recognition is a subject of informatics. There are well established standards for formulation of different problems in this subject. However, the formulation of the problem by the author does not correspond to the standards. In the paper a problem of recognition is considered where signals are elements of a Hilbert space, but description of classes is confusing. Assumptions about classes are hidden in a mess of terms of fuzzy logic, probabilities, and quantum theory. Faults in formulation of the problem are serious enough to be the only reason for rejection. But they are not the only reason. The intended results of the paper are not clearly formulated. What kind of applied problems are aimed by means of the considered methods? Where the considered methods are expected to be competitive? Formulae manipulations seem trivial and not of interest for the audience of Informatica. At least there is no clear exposition of any new theoretical results. English of the paper needs an extensive editing; it is too much of sweating for a reviewer to list all errors and inaccuracies. Так что всё тривиально. А если постановка задачи не удовлетворяют каким-то мне непонятным стандартам (, то это Fault. Логика у рецензента просто поразительная, но явна не математическая. Хоть бы указал, где написаны эти стандарты читать.

bne: Мне идея показалась интуитивно прозрачной и разумной А что удалось формализовать можно только приветствовать Ну, а что не сильно им нравятся конкуренты и они содержанию предпочитают форму меня не удивляет (хотя мне неприятно обсуждать эту тему< поскольку слишком многие политиканы на этом спекулируют)

grisania: Спасибо за поддержку, но очень обидно все равно. Я же просто дал некоторое обоснавание методу главных компонент. Я смотрю, что на этом форуме интересуются РСА. Также дал обоснование методу подпространств - Method CLAFIC, используя логику подпространств или фильтров. Так уж получилась, что эта логика давно используется в квантовой механике - КМ . В КМ фильтры пропускают таинственную квантовую вероятность. Как говорил Эйнштейн "Бог не играет в кости". Обычно фильтры строят в спектральной области. Я вообще сменю свой ник на ник "grisania", мне он больше нравится.

bne: Ограничения на число ников наверняка есть, но пока до них явно далеко ;-) Пишите о своих работах - это интересно Авось и подберем здесь неплохую компанию Кстати, можно ссылки размесиить на тех же Вапника, Червоненкиса, Моттля , Миркина и прочих эмигрантов

grisania: Было бы интересно делать совместные статьи, какие-то проекты. Вон как програмеры клепают свободный софт. Спасибо амерам за их arXiv. Наука должна меняться в сторону открытости, это как средние века стали рушиться Цеха мастеров, рыцарское сословия. Когда я учился в аспирантуре, была такая шутка, руководитель имеет право 1-ой брачной ночи с невестой своего аспиранта

bne: На сходные темы полно рассуждений и споров выше (в разделе методология) Что касается совместной деятельности - нужны продуманные организационные формы при которых сохраняется интерес всех сторон Про право на аспиранта тема известная Так западники временами имеют наших математиков и физиков А порой и геофизиков

bne: Логические аппроксимации, лапласовы оценки и корреляционная логика #11 ноябрь 2005 Г. Н. Зверев, д-р техн. наук, Уфимский государственный технический университет Логические аппроксимации, лапласовы оценки и корреляционная логика Точные решения дискретно-логических задач с полной априорикой формализма частотной логики заменяются в реальных информационных ситуациях логическими приближениями, основанными на лапласовых оценках высших моментов по известным низшим, снижении размерности и экспоненциальной сложности логических задач. Современные информационные технологии и их программно-аппаратные средства основаны на формализме классической логики с двоичной шкалой различимости {да, нет} и истинности {истина, ложь} ? {1, 0}. Однако реальные информационные процессы оперируют искаженными, неполными, противоречивыми фактическими и априорными данными, которым вряд ли возможно приписать однозначную оценку 0 или 1 двоичной шкалы. В таких ситуациях более адекватным описанием является формализм частотной логики [1—3], в котором двоичная истинность заменяется частотной истинностью, а логические связи двоичных признаков, свойств и предикатов представлены различными частотными связями. Последние в пределе могут сколь угодно приближаться к строгим логическим оценкам и зависимостям. Переход от предельных, недостижимых оценок истинности в двоичной шкале {1, 0} к числовым оценкам истинности в открытом интервале (0, 1) позволяет построить логические аппроксимации, которые учитывают искажения и неопределенности фактов и априорики. Целью данной статьи является построение оценок истинности логических методов и погрешностей истинности в корреляционном приближении. .... Список литературы 1. Зверев Г. Н. Оценка надежности и оптимизация качественной интерпретации // Техника и технология геофизических исследований скважин. Уфа: Изд. БашНИПИнефть, 1979. С. 145-155. 2. Зверев Г. Н. Основания теоретической информатики. Разд. 1—7 Уфа: УГАТУ, 1997. 97 с. 3. Зверев Г. Н. Точные и аппроксимационные логики в машинных рассуждениях // Тр. V нац. конференции по искусственному интеллекту. Т. 1. Казань. 1996. С. 62—66. ИНФОРМАЦИОННЫЕ ТЕХНОЛОГИИ, № 2, 1999 ИНТЕЛЛЕКТУАЛЬНЫЕ СИСТЕМЫ Ключевые слова: Дискретная логика, частотная логика, методы корреляции, частотная связность признаков, корреляция моментов высших порядков. Полный текст статьи по ссылке

grisania: Зачем геофизикам моя статья? Она больше теоритическая, чем прикладная. Ваша тема и называется "Статистические задачи и алгоритмы", но, как я понимаю, без теоритических изысков. Я сам программирую на всех языках, неплохо владею МатЛабом. Я на самом деле преподаю программирование Могу посчитать на геофизических данных какую нибудь задачу распознавания с применением Energy Discriminant Analysis. Я уже писал, что я сам применял этот метод для распознавания фонем. Пришлось написать кучу скриптов на МатЛабе.

bne: 1) Не стоит думатm, что все геофизики совсем тупые. Те же программы Бонгарда быстрее всех и лучше всех раскручивались на каротаже (Губерманом) Когда Вапник с Червоненкисом были еще в ИПУ он активно сотрудничали с ЦГЭ Потом несколько лет это делал Моттль (в связи с корреляцией разрезов и иными применениями задач о разладке в марковских полях) Так что деталей можем не понимать, но суть обычно схватываем 2) Нечеткие множества (как и нейронные сети от SOM до RBF) также не прошли мимо геофизики - полно публикаций 3) По моему опыту успех в решении задач прикладной статистики состоит в умении экстрагировать ограничения на решения (априорную информацию) и встроить ее в алгоритм максимально естественным образом Хорошие постановки задач обычно это позволяют 4) Задача с фонемами судя по тому что я на сей счет слышал весьма противная - наверное немало пришлось с ней возиться< Но это подтверждает качество алгоритма 5) На мой взгляд Вы можете найти себе здесь заинтересованных заказчиков или под задачи или под алгоритмическую поддержку 6) matlab - хорошая среда, но IMHO научная Реально опытов продажи софта на ней на вертикальном рынке я не знаю Иногда можно перелицовывать тексты, а потом прописывать в системы Наверное это предмет 7) Не знаю как другие фирмы, но ПАНГЕЯ (в сейсмической части) имеет такой опыт Решается тестовая задача и если она лучше или соизмеримо с применяемыми или технологична - заключается договор на разработку софта Я не менеджер и потому детали (авторские права и стоимости) обсуждать не могу Это индивидуально

viking23: матлаб включает компилятор языка C++. но что я вам скажу писать на нем намного быстрее. недавно закончил писать утилиту для объединения разных петрофизических результатов в excel файлах. Нудное занятие. И пока писал на C# вспоминал, как легко циклы было делать на MatLab. Единственное что меня не удовлетворяет в последнем, то что нет поддержки fla -флеш. А это сейчас будущее. В частности для создании активного приложения в презентации, по попросту - видео, но без встраивания видео роликов(не у всех есть кодаки), flash не заменим. Единственный аналог анимированый gif требует наличия MS office 2003, для внедрения в документ презентации, и является не управляемым (запускается сразу при показе данного слайда с картинкой gif). Таким образом пока нельзя качественно воспроизвести данные, изменяющиеся во времени. А что это за результаты работы которые нельзя показать?))

bne: Известная книга "Классификация и кластер" (сборник переводов) Лежит по ссылке: http://lib.socio.msu.ru/l/library?e=d-000-00---001ucheb--00-0-0-0prompt-10---4------0-1l--1-ru-50---20-help---00031-001-1-0windowsZz-1251-00&a=d&c=01ucheb&cl=CL1&d=HASH0101541ec2bbe3c2964e4d34 К сожалению, все в HTML

bne: Конференции «Математические методы распознавания образов» http://www.mmro.ru/proccoll.php

bne: T. Yu and D.Wilkinson: A Co-Evolutionary Fuzzy System for Reservoir Well Logs Interpretation, Studies in Computational Intelligence (SCI) 88, 199–218 (2008) www.springerlink.com Springer-Verlag Berlin Heidelberg 2008 Забавная статья по применению генетических алгоритмов

bne: Оригиналы у меня отсутствуют A Markov-Gauss algorithm for blocking well logs. Authors: Chouinard, Paul N. Publication: Geophysics, vol. 53, issue 8, p. 1118. Segmentation of well logs by maximum-likelihood estimation Журнал Mathematical Geology Номер Volume 22, Number 7 / Октябрь 1990 г. Страницы 853-869

bne: проблемы (когда-то писал про это раздел в справочник) Старые алгоритмы (Андерссон или Родионов или Губерман) основаны на сходных идеях (отличия фоновых вариаций от базовых), но обычно не сильно помогают Часто дополнительные фокусы в плохой увязке кривых между собой (что особенно существенно при нескольких спуско-подъемах) Тут нужно применять приемы близкие к тем которые применяют для обработки наклономера или корреляции (пор сути - динамическое программирование)

bne: Еще одна классика - Губерман с его КЧП (это также начало 70-х - до перестройки) КЧП означает К ЧЕРТУ ПОДРОБНОСТИ По сути, речь идет о выборе порога огрубления при выделении остова при котором все не начинает размазываться Губерман и его сотрудники (Овчинникова и другие) применял это не только при расчленении каротажа для корреляции (потом его разработку и сотрудников подхватил Гутман, в связи с корреляцией разрезов, и (как мне кажется незаслуженно принизил), но и для обработки флюрографии Нахваливал он метод основательно (в принципе может так и надо, но меня шокировало по молодости) В принципе сходство с КЧП есть, вопрос в нащупывании порога и стратегии этого нащупывания (можно ли устраивать качели и как останавливаться) Думаю, что окультуривание подхода состоит в правильном использованиии 1) модели, 2) байеса и априорных вероятностей, 3) аналога EM или самого EM.

bne: В принципе соображения Губермана в КЧП IMHO восходят к идея Беллмана об инвариантном погружении Задаются формальные параметры широкого класса сред и потом выбираются критичные В ранней работе Губерман со вкусом писал про овражность (в смысле выделения для подобного параметра диапазонов) и ссылался на метод оврагов Гелфанда-Цетлина и на задачу о выборе энергии при работе пула радиостанций Сама по себе идея Беллмана и ее применение симпатичны, а лирика вокруг IMHO скорее эвристична и в большей мере по части демонстрации статуса

bne: И все же пробный счет IMHO показывает, что аналоги медианной фильтрации на малой базе работают лучше Другое дело, что параметры для них можно подбирать на большой базе

bne: Можно, конечно, ставить задачу как оптимизационную (можно и критерий выписать, понятно, что с константами отражающими особенности разреза) В принципе (в одном измерении) она и сепарабельная Но многоэкстремальная Отсюда и соответствующие методы (или динамическое программирование или игры в GA или аннилинг, пусть и с локальным спуском) Но можно попробовать обойти ситуацию и со стороны EM алгоритмов

bne: A History of Regression and Related Model-Fitting in the Earth Sciences (1636?-2000) Natural Resources Research Springer Netherlands ISSN 1520-7439 1573-8981 Volume 10, Number 4 / Декабрь 2001 г. DOI 10.1023/A:1013928826796 Страницы 241-286 A History of Regression and Related Model-Fitting in the Earth Sciences (1636?-2000) Richard J. Howarth1 (1) Department of Geological Sciences, University College London, Gower Street, London, WC1E 6BT, UK Abstract The (statistical) modeling of the behavior of a dependent variate as a function of one or more predictors provides examples of model-fitting which span the development of the earth sciences from the 17th Century to the present. The historical development of these methods and their subsequent application is reviewed. Bond's predictions (c. 1636 and 1668) of change in the magnetic declination at London may be the earliest attempt to fit such models to geophysical data. Following publication of Newton's theory of gravitation in 1726, analysis of data on the length of a 1° meridian arc, and the length of a pendulum beating seconds, as a function of sin2(latitude), was used to determine the ellipticity of the oblate spheroid defining the Figure of the Earth. The pioneering computational methods of Mayer in 1750, Boscovich in 1755, and Lambert in 1765, and the subsequent independent discoveries of the principle of least squares by Gauss in 1799, Legendre in 1805, and Adrain in 1808, and its later substantiation on the basis of probability theory by Gauss in 1809 were all applied to the analysis of such geodetic and geophysical data. Notable later applications include: the geomagnetic survey of Ireland by Lloyd, Sabine, and Ross in 1836, Gauss's model of the terrestrial magnetic field in 1838, and Airy's 1845 analysis of the residuals from a fit to pendulum lengths, from which he recognized the anomalous character of measurements of gravitational force which had been made on islands. In the early 20th Century applications to geological topics proliferated, but the computational burden effectively held back applications of multivariate analysis. Following World War II, the arrival of digital computers in universities in the 1950s facilitated computation, and fitting linear or polynomial models as a function of geographic coordinates, trend surface analysis, became popular during the 1950–60s. The inception of geostatistics in France at this time by Matheron had its roots in meeting the evident need for improved estimators in spatial interpolation. Technical advances in regression analysis during the 1970s embraced the development of regression diagnostics and consequent attention to outliers; the recognition of problems caused by correlated predictors, and the subsequent introduction of ridge regression to overcome them; and techniques for fitting errors-in-variables and mixture models. Improvements in computational power have enabled ever more computer-intensive methods to be applied. These include algorithms which are robust in the presence of outliers, for example Rousseeuw's 1984 Least Median Squares; nonparametric smoothing methods, such as kernel-functions, splines and Cleveland's 1979 LOcally WEighted Scatterplot Smoother (LOWESS); and the Classification and Regression Tree (CART) technique of Breiman and others in 1984. Despite a continuing improvement in the rate of technology-transfer from the statistical to the earth-science community, despite an abrupt drop to a time-lag of about 10 years following the introduction of digital computers, these more recent developments are only just beginning to penetrate beyond the research community of earth scientists. Examples of applications to problem-solving in the earth sciences are given. History of geology - history of geophysics - CART - errors-in-variables models - geostatistics - kernel regression - kriging - least squares - logistic regression - LOWESS - mixture models - multiple linear regression - nonlinear regression - Poisson regression - regression analysis - ridge regression - robust regression - splines - trend-surface analysis -------------------------------------------------------------------------------- Richard J. Howarth Email: r.howarth@ucl.ac.uk REFERENCES Abbe, C., 1871, A historical note on the method of least squares: Am. Jour. Science, Ser. 3, v. 1, no. 6, p. 411–415. Abbey, S., 1986, A study of robust estimators: Geostandards Newsletter, v. 10, no. 2, p. 159–168. Abbey, S., 1992, Evaluation and application of reference materials for the analysis of rocks and minerals: Chem. Geology, v. 95, no. 1-2, p. 123–130. Adcock, R. J., 1878, A problem in least squares: The Analyst (Des Moines, Iowa), v. 5, p. 53–54. Adrain, R., 1808, Research concerning the probabilities of the errors which happen in making observations: The Analyst or Mathematical Mus. Philadelphia, v. 1, no. 4, p. 93–109. Adrain, R., 1818, Investigation of the Figure of the Earth, and of the gravity in different latitudes:Am.Philosophical Soc.Trans., n.s., v. 1, p. 119–135. Agocs, W. B., 1951, Least squares residual anomaly determination: Geophysics, v. 16, no. 4, p. 686–696. Agterberg, F. P., 1974, Automatic contouring of geological maps to detect target areas for mineral exploration: Math. Geology, v. 6, no. 4, p. 373–395. Agterberg, F. P., 1988, Quality of time scales-a statistical appraisal, in Merriam, D. F., ed., Current Trends in Geomathematics: Plenum, New York, p. 57–103. Agterberg, F. P., 1990, Automated stratigraphic correlation: Elsevier, Amsterdam, 424 p. Agterberg, F. P., 1994, Estimation of the Mesozoic geological time scale: Math. Geology, v. 26, no. 7, p. 857–876. Agterberg, F. P., and Cabilio, P., 1969, Two-stage least-squares model for the relationship between mappable geological variables: Math. Geology, v. 6, no. 4, p. 37–53. Agterberg, F. P., and Chung, C. F., 1975, A computer-program for polynomial trend-surface analysis: Geol. Survey Canada Paper, v. 75-21, p. 1–16. Agterberg, F. P., and Robinson, S. C., 1972, Mathematical problems in geology [with discussion]: Intern. Statistical Inst. Bull. v. 44, no. for 1971, Book 1, p. 567–595. Airy, G. B., 1845, Figure of the Earth, in Smedley, E., Rose, H. J., and Rose, H. J., eds., Encyclopaedia Metropolitana, v. 5. Fellowes and Rivington, London, p. 175–210. Airy, G. B., 1861, On the algebraical and numerical theory of errors of observations and the combination of observations: Macmillan, Cambridge, p. 103 C (loose undated supplement) p. 107-118. Aitchison, J., 1986, The statistical analysis of compositional data: Chapman and Hall, London, 416 p. Albarede, F., and Provost, A., 1977, Petrological and geochemical mass-balance equations: an algorithm for least-square fitting and general error analysis: Computers & Geosciences, v. 3, no. 2, p. 309–326. Alkins, W. E., 1920, Morphogenesis of Brachiopoda. I. Reticularia lineata (Martin), Carboniferous Limestone: Mem. and Proc. Manchester Lit. and Phil. Soc., v. 64, no. 2, p. 1–11. Alldredge, J. R., and Alldredge, N. G., 1978, Geostatistics: a bibliography: Intern. Stat. Review, v. 46, no. 1, p. 77–88. Allen, P., and Krumbein, W. C., 1962, Secondary trend components in the top Ashdown pebble bed, a case history: Jour. Geology, v. 70, no. 5, p. 507–538. Anderson, T. W., Gupta, S. D., and Styan, G. P. H., 1972, A bibliography of multivariate statistical analysis: Oliver and Boyd, Edinburgh, 642 p. Andrews, D. F., 1974, A robust method for multiple linear regression: Technometrics, v. 16, no. 4, p. 523–531. Anonymous, 1821, Dissertation sur la recherche du milieu le plus probable, entre les r ´ esultats de plusieurs observations ou experi´ ences: Annales de Mathematiques, v. 12, p. 181–204. Anonymous, 1838, Introduction: Jour. Stat. Soc. London, v. 1, no. 1, p. 1–5. Anscombe, F. J., 1961, Examination of residuals, in Neyman, J., ed., Proc. Fourth Berkely Symp. Mathematical Statistics and Probability, 4 vols: Univ. California Press, Berkeley, v. 1, p. 1–36. Anscombe, F. J., and Tukey, J.W., 1963, The examination and analysis of residuals: Technometrics, v. 5, no. 1, p. 141–160. Anselone, P. M., and Laurent, P. J., 1968, A general method for the construction of interpolating or smoothing spline functions: Numerische Mathematik, v. 12, p. 66–82. Baird, A. K., McIntyre, D. B., and Welday, E. E., 1967, Geochemical and structural studies in batholithic rocks of southern California: Part II, Sampling of the Rattlesnake Mountain Pluton for chemical composition, variability, and trend analysis: Geol. Soc. America Bull., v. 78, no. 2, p. 191–222. Bannister, F. A., 1931, A relation between the density and refractive index of silicate glasses, with application to the determination of imitation gemstones: Mineral. Magazine, v. 22, no. 126, p. 136–154. Barlow, P., 1833, On the present situation of the magnetic lines of equal variation and their changes on the terrestrial surface: Royal Society Philosophical Trans., v. 123, p. 667–673. Bartlett, M. S., 1949, Fitting a straight line when both variables are subject to error: Biometrics, v. 5, no. 3, p. 207–212. Barton, 1929, Control and adjustment of surveys with the magnetometer and torsion balance: Am. Assoc. Petroleum Geologists Bull., v. 13, no. 9, p. 1163–1186. Bauer, L. A., 1913, The instruments and methods of research: Proc.Washington Philosophical Soc., v. 15, no. for 1906-1910, p. 103–126. Belonin, M. D., and Zhukov, I. M., 1970, Geometrical properties of the subsurface of the Alekseevka Uplift in the Kuibyshev District, in Romanova, M. A., and Sarmanov, O. V., eds., Topics in Mathematical Geology: Plenum Press, New York, p. 186–199. Belsley, D. A., Kuh, E., and Welsch, R., 1980, Regression diagnostics: identifying influential data and sources of collinearity: John Wiley & Sons, New York, 292 p. Berkson, J., 1944, Application of the logistic function to bioassay: Jour. Am. Stat. Assoc., v. 39, no. 227, p. 357–365. Bessel, F. W., 1841, Determination of the axes of the elliptic spheroid of revolution which most nearly corresponds with the existing measurements of arcs of the meridian: Taylor's Scientific Mem., v. 2, no. 7, p. 387–400. Bessel, F. W., and Baeyer, J. J., (1838) 1876, Gradmessung in Ost-Preussen und ihre Verbindung mit preussische und russische Dreicksketten, in Englemann, R., ed., Abhandlungen von Friedrich Wilhelm Bessel. III. Geodaesie, Physik, Verschiedenes, Literatur: Englemann, Berlin, p. 62–138. Bickel, P. J., 1965,Onsome robust estimates of location: Ann. Math. Statistics, v. 36, no. 3, p. 847–858. Biot, J. B., 1816, Traité de physique expérimentale et mathématique:, v. 3, Deterville, Paris, 516 p. Biot, J. B., 1829, Mémoire sur la Figure de la Terre: Mémoires de l'Académie Royale des Sciences de l'Institut de France, Paris, v. 8, p. 1–56. Birkhoff, G., 1980, Computing developments 1935-1955, as seen from Cambridge, U.S.A., in Metropolis, N., Howlett, J., and Rota, G.-C., eds., A History of Computing in the Twentieth Century: Academic Press, New York, p. 21–30. Bond, H., 1668, The variations of the magnetic needle predicted for many years following: Royal Society London, Philosophical Trans. v. 3, p. 789–790. Bond, H., 1673, Untitled broadsheet: British Library, Add.ms. 4393, London, ff. 38. Bond, H., 1674, The use of the Inclynatorie Magneticall Needle to find the longitude by the use of the Carolina tables: British Library, London, Add. ms. 4393, ff. 7-20. Bond, H., 1676, The longitude found: or, a treatise shewing an easie and speedy way, as well by night as by day, to find the longitude, having but the latitude of the place, and the inclination of the Magnetical Inclinatorie Needle: Godbid, London, 65 p. Bowditch, N., 1829-39, Méchanique Céleste by the Marquis de La Place translated with a commentary by Nathaniel Bowditch: 4 vols., Hilliard, Gray, Little andWilkins, Boston, 746 p. (1829); 990 p. (1832); 910 p. (1834); 1018 p. (1839). Bowen, N. L., 1928, The evolution of the igneous rocks: Princeton Univ. Press, Princeton, New Jersey, 332 p. Box, G. E. P., 1953, Non-normality and tests on variances: Biometrika, v. 40, no. 3-4, p. 318–335. Breiman, L., Friedman, J. H., Olshen, R., and Stone, C. J., 1984, Classification and regression trees:Wadsworth, Belmont, California, 358 p. Brinkley, S. R., 1946, Note on the conditions of equilibrium for systems of many constituents: Jour. Chemical Physics, v. 14, p. 563–566. Brinkmann, R., 1929, Statistisch-biostratigraphische Untersuchungen an mitteljurassischen Ammonitenüber Artbegriff und Stammesentwicklung: Abhandlungen der Gesellschaft der Wissenschaften zu Göttingen. Mathematisch-physikalische Klasse, f., v. 13, no. 3, p. 1–249. Bullard, E. C., 1936, Gravity measurements in East Africa: Royal Soc. Philosophical Trans., v. A235, no. 757, p. 445–531. Bullard, E. C., 1939, Heat flow in South Africa: Royal Soc. London Philosophical Trans., v. A173, no. 955, p. 474–502. Bullard, E. C., Gaskell, T. F., Harland, W. B., and Kerr-Grant, C., 1940, Seismic investigations on the Palaeozoic floor of East England: Royal Soc. London Philosophical Trans., v. A239, no. 800, p. 29–93. Bumstead, H. A., and Wheeler, L. P., 1904, On the properties of a radio active gas found in the soil and water near New Haven: Am. Jour. Science. Ser. 4, v. 17, no. 98, p. 97–111. Burma, B. H., 1949, Studies in quantitative paleontology. II. Multivariate analysis-a new analytical tool for paleontology and geology: Jour. Paleontology, v. 23, no. 1, p. 95–103. Butler, R. D., 1933, Immersion liquids of intermediate refraction (1.450-1.630): Am. Mineralogist, v. 18, no. 9, p. 386–401. Carroll, R. J., and Ruppert, D., 1996, The use and misuse of orthogonal regression in linear errors-in-variables models: Am. Statistician, v. 50, no. 1, p. 1–6. Carroll, R. J., and Spiegelmann, C. H., 1992, Diagnostics for nonlinearity and heteroscedasticity in errors-in-variables regression: Technometrics, v. 32, no. 2, p. 186–196. Chayes, F., 1970, On deciding whether trend surfaces of progressively higher order are meaningful: Geol. Soc. America Bull., v. 81, no. 4, p. 1273–1278. Chebyshev, P. L., (1855) 1899, Sur les fractions continues; orig. in: Journal de Mathématiques Pures et Appliqués, v. 3, no. 15; reprinted in Markoff, A., and Sonin, N., eds., 1899, Oeuvres de P. L. Tchebychev: Glasounof and Eggers, St. Petersberg, v. 1, p. 203–230. Chebyshev, P. L., (1859) 1899, Sur l'interpolation par laméthode des moindres carrés; orig. in M´emoires de l'Académie Imperiale des sciences de St. Petérsbourg. ser. 7. v. 1, no. 5; reprinted in Markoff, A., and Sonin, N., eds., Oeuvres de P. L. Tchebychev; Glasounof and Eggers, St. Petersberg, v. 1, p. 473–498. Chebyshev, P. L., (1864) 1899, Sur l'interpolation; orig. in: Procédé de l'Académie imperiale des Sciences de St. Petérsbourg, v. 4, no. 5; reprinted in Markoff, A., and Sonin,N., eds., 1899, Oeuvres de P. L. Tchebychev: Glasounof and Eggers, St. Petersberg, v. 1, p. 541–560. Chebyshev, P. L., (1875) 1907, Sur l'interpolation des valeurs équidistantes; orig. in Procédé de l'Académie imperiale des Sciences de St. Pet´ersbourg, v. 25, no. 5; reprinted in Markoff, A., and Sonin,N., eds., Oeuvres deP.L. Tchebychev: Glasounof and Eggers, St. Petersberg, v. 2, p. 219-24. Cheng, C.-L., and Schneeweiss, H., 1998, Polynomial regression with errors in the variables: Jour. Royal Stat. Soc., v. B60, no. 1, p. 189–199. Chork, C. Y., 1991, An assessment of least median of squares regression in exploration geochemistry: Jour. Geochem. Exploration, v. 41, no. 3, p. 325–340. Chung, C. F., and Agterberg, F. P., 1980, Regression models for estimating mineral resources from geological map data: Math. Geology, v. 12, no. 5, p. 473–488. Clark, I., 1977a, SNARK-A four-dimensional trend-surface program: Computers & Geosciences, v. 3, no. 2, p. 283–308. Clark, I., 1977b, ROKE, a computer program for nonlinear leastsquares decomposition of mixtures of distributions: Computers & Geosciences, v. 3, no. 2, p. 245–256. Clark, M.W., 1977, GETHEN: a computer program for the decomposition of mixtures of two normal distributions by the method of moments: Computers & Geosciences, v. 3, no. 2, p. 257–267. Cleveland, W. S., 1979, Robust locally weighted regression and smoothing scatterplots: Jour. Am. Stat. Assoc., v. 74, no. 368, p. 829–836. Cochran, W. G., 1940, The analysis of variance when experimental errors follow the poisson or binomial law: Ann. Math. Statistics, v. 11, no. 3, p. 335–347. Cohen, R. E., 1986, Thermodynamic solution properties of aluminous clinopyroxenes: nonlinear least squares refinements: Geochem. et Cosmochim. Acta, v. 50, no. 4, p. 563–575. Cole, A. J., ed., 1969, An iterative approach to the fitting of trend surfaces. Kansas Geol. Survey, Computer Contrib. 37, 27 p. Connor, J. J., and Miesch, A. T., 1964, Analysis of geochemical prospecting data from the Rocky Range, Beaver County, Utah: U. S. Geol. Survey Prof. Paper, v. 475D, p. 79–83. Cook, R. D., 1977, Detection of influential observations in linear regression: Technometrics, v. 19, no. 1, p. 15–17. Cook, R. D., and Weisberg, S., 1982, Residuals and influence in regression: Chapman and Hall, London, 227 p. Cooley, R. L., 1982, Incorporation of prior information on parameters into nonlinear regression groundwater flow models: I, Theory: Water Resources Research, v. 18, no. 4, p. 965–976. Cooley, R. L., 1983, Incorporation of prior information on parameters into nonlinear regression groundwater flow models: II, Applications: Water Resources Research, v. 19, no. 3, p. 662–676. Cotes, R., 1722, Aestimatio errorum in mixta mathesi, per variationes partium trianguli plani et spherici, in Smith, R., ed., Harmonia Mensuram, sive Analysis et Synthesis. II. Opuscula Mathematica: Cambridge, p. 1-22. Craven, P., and Wahba, G., 1979, Smoothing noisy data with spline functions: estimating the correct degree of smoothing by the method of generalized cross-validation: Numerische Mathematik, v. 31, p. 377–403. Cressie, N., 1990, The origins of Kriging: Math. Geology, v. 22, no. 3, p. 239–252. Cressie, N., 1991, Statistics for spatial data: John Wiley & Sons, New York, 900 p. Cressie, N., and Hawkins, D. C., 1980. Robust estimation of the variogram: Math. Geology, v. 12, no. 2, p. 115–125. Curry, H. B., and Schoenberg, I. J., 1966, On Polya frequency functions. IV: Jour. d'Analyse Math´ematique, v. 17, p. 71–107. Dahlberg, E. C., 1969, Use of model for relating geochemical prospecting data to geologic attributes of a region, South Mountain, Pennsylvania: Quart. Jour. Colorado Sch. Mines, v. 64, no. 1, p. 195–216. Dahlberg, E. C., and Griffiths, J. C., 1967, Defining gradients in a sample of sedimentary rock: Sedimentology, v. 8, no. 4, p. 291–309. David, M., 1977, Geostatistical ore reserve estimation: Elsevier, Amsterdam, 364 p. Davis, D.W., 1982, Optimal linear regression and error estimation applied to U-Pb data: Can. Jour. Earth Sciences, v. 19, no. 11, p. 2141–2149. Davis, J.C., 1973, Statistics and data analysis in geology: JohnWiley & Sons, New York, 550 p. Dawson, K. R., and Whitten, E. H. T., 1962, The quantitative mineralogical composition and variation of the Lacorne, La Motte, and Preissac Granitic Complex, Quebec, Canada: Jour. Petrology, v. 3, no. 1, p. 1–37. DeGeoffroy, J. E., and Wignall, T. K., 1971,Aprobabilistic appraisal of mineral resources in a portion of the Grenville Province of the Canadian Shield: Econ. Geology, v. 66, no. 3, p. 466–479. DeLury, D. B., 1950, Values and integrals of the orthogonal polynomials up to n D 26: Univ. Toronto Press, Toronto, 33 p. Doveton, J. H., and Parsley, A. J., 1970, Experimental evaluation of trend surface distortions induced by inadequate data-point distributions: Trans. Instit. Mining and Metallurgy, London, v. B79, no. 4, p. 197–207. Draper, N. R., and Smith, H., 1966, Applied regression analysis: John Wiley & Sons, New York, 407 p. Draper, N. R., and Smith, H., 1967, Selecting the best regression equation: Kansas Geol. Survey, Computer Contrib. 12, p. 1–3. Dubrule, O., 1984, Comparing splines and kriging: Computers & Geosciences, v. 10, no. 2-3, p. 327–338. Dunnington, G. W., 1955, Carl Friedrich Gauss. Titan of Science: Hafner, New York, 479 p. Dwyer, P. S., and Waugh, F. V., 1953, On errors in matrix inversion: Jour. Am. Stat. Assoc., v. 48, no. 262, p. 289–319; Corrigendum no. 264, p. 911-912. Eckhardt, D. A., Flipse, W. J., and Oaksford, E. T., 1989, Relation between land use and ground-water quality in the upper glacial aquifer in Nassau and Suffolk Counties, Long Island, NY: U.S. Geol. SurveyWater Resources Invest. Rept. no. 86-4142, p. 1–26. Edgeworth, F. Y., 1883, The method of least squares: Philosophical Magazine, Ser. 5, v. 16, no. 101, p. 360–375. Edgeworth, F. Y., 1887, On observations relating to several quantities: Hermathena, Dublin, v. 6, p. 279–285. Edgeworth, F. Y., 1892, Correlated averages: Philosophical Magazine, Ser. 5., v. 34, no. 207, p. 190–204. Efroymson, M. A., 1960, Multiple regression analysis, in Ralston, A., and Wilf, H. S., eds., Mathematical Methods for Digital Computers: John Wiley & Sons, New York, p. 191–203. Eisenhart, C., 1961, Boscovich and the combination of observations, in Whyte, L. L., ed., Roger Joseph Boscovich, S. J., F. R. S., 1711-1787. Studies of his life and work on the 250th anniversary of his birth: Allen and Unwin, London, p. 200–212. Eisenhart, C., ed., 1978, Carl Friedrich Gauss, in Kruskal, W. H., and Tanur, J. M., eds., International Encyclopaedia of Statistics, 1: Free Press, New York, p. 378–386. Encke, J. F., 1834, Ueber die Methode der kleinsten Quadrate. I.: Berliner astronomische Jahrbuch, no. for 1834, p. 249–312. Encke, J. F., 1835, Ueber die Methode der kleinsten Quadrate. II.: Berliner astronomische Jahrbuch, no. for 1835, p. 253–320. Encke, J. F., 1836, Ueber die Methode der kleinsten Quadrate. III.: Berliner astronomische Jahrbuch, no. for 1836, p. 253–308. Encke, J. F., 1841,Onthe method of least squares:Taylor's Scientific Mem., v. 2, no. 7, p. 317–369. Eskola, P., 1922, The mineral facies of rocks: Norsk geologisk Tidsskrift, v. 6, no. 1, p. 143–194. Esler, J. E., Smith, P. E., and Davis, J. C., 1968, KWIKR8-A FORTRAN IV program for multiple regression and geologic trend analysis: Kansas Geol. Survey, Computer Contrib. 28, 31 p. Evans, G., Howarth, R. J., and Nombela, M. A., in prep., Metals in the sediments of the Ensenada de San Simòn (inner Ría de Vigo), Galicia, N.W. Spain. Everett, R. R., 1980, WHIRLWIND, in Metropolis, N., Howlett, J., and Rota, G.-C., eds.,AHistory of Computing in theTwentieth Century: Academic Press, New York, p. 365–384. Fischer, I., 1975, The figure of the Earth-changes in concepts: Geophys. Surveys, v. 2, no. 1, p. 3–54. Fisher, R. A., 1922a, The goodness of fit of regression formulae and the distribution of regression coefficients: Jour. Royal Stat. Soc., v. 85, no. 4, p. 597–612. Fisher, R. A., 1922b, On the mathematical foundations of theoretical statistics: Philosophical Trans. Royal Society, London, v. A222, p. 309–368. Fisher, R. A., 1925, The application of Student's distribution: Metron, v. 5, no. 3, p. 90–104. Forbes, J.D., 1840, Account of some experiments made in different parts of Europe, on terrestrial magnetic intensity, particularly with reference to the effect of height: Royal Society Edinburgh Trans., v. 14, pt. 1, p. 1–29. Forrest, D.W., 1974,Francis Galton.The life and work of aVictorian genius. Elek, London, 340 p. Fox, W. T., 1967, FORTRAN IV program for vector trend analysis of directional data: Kansas Geol. Survey, Computer Contr. 11, 36 p. Freund, R. J., 1963,Awarning of roundoff errors in regression: The Am. Statistician, v. 17, no. 5, p. 13–15. Furnival, G. M., 1971, All possible regressions with less computation: Technometrics, v. 13, no. 2, p. 403–408. Furnival, G. M., and Wilson, R. W., 1974, Regression by leaps and bounds: Technometrics, v. 16, no. 4, p. 499–511. Galloway, T., 1846, On the application of the method of least squares to the determination of the most probable errors in a portion of the Ordnance Survey of England: Royal Astron. Soc., London Mem., v. 15, p. 23–69. Galton, F., 1875, Statistics by intercomparison, with remarks on the law of frequency of error: Philosophical Magazine, Ser 4, v. 49, no. 322, p. 33–46. Galton, F., 1877, Typical laws of heredity: Nature, v. 15, no. 388, p. 492–495, no. 389, p. 512-514, no. 390, 532-533. Galton, F., 1888, Co-relations and their measurement, chiefly from anthropometric data: Proc. Royal Society, London, v. 45, no. 274, p. 135–145. Galton, F., 1889, Natural inheritance: Macmillan, London, 259 p. Ganse, R. A., Amemiya, Y., and Fuller, W. A., 1983, Prediction when both variables are subject to error, with application to earthquake magnitudes: Jour. Am. Stat. Assoc., v. 78, no. 4, p. 761–765. Garrett, R. G., 1983, Opportunities for the 80s: Math. Geology, v. 15, no. 2, p. 385–398. Garrett, R. G., Goss, T. I., and Poirier, P. R., 1982, Multivariate outlier detection-an application to robust regression in the earth sciences (abst.): Joint Stat. Meetings Am. Stat. Assoc. (Cincinnati, Ohio) p. 101. Gastwirth, J., 1966, On robust procedures: Jour. Am. Stat. Assoc., v. 61, no. 316, p. 929–948. Gauss, C. F., 1809, Determinatio orbitae observationibus quotcunque maxime satisfacientis, in Theoria Motus Corporum Coelestium in Sectionibus Conicis Solem Ambientium: Perthes and Besser, Hamburg, p. 205–224; English transl., Theory of the motion of the heavenly bodies moving about the sun in conic sections [selected extracts], C. H. Davis transl. (1857), in Shapley, H., and Howarth, H. E., eds., 1929, A Source Book in Astronomy: McGraw-Hill, New York, p. 183-195. Gauss, C. F., 1811, Disquisito de elementis ellipticis Palladis ex oppositionibus 1803, 1804, 1805, 1807, 1808, 1809: Commentationes societas regiae scientiarum Göttingensis recentiores, v. 1, p. 1–26. Gauss, C. F., 1823, Theoria combinationis observationum erroribus minimis obnoxiae: Commentationes societas regiae scientiarum Göttingensis recentiores, v. 5, no. for 1821-3, p. 33–90. Gauss, C. F., 1832, Intensitas vis magneticae terrestris ad mensuram absolutam revocata: Göttingische gelehrte Anzeigen, v. 205, no. 7, p. 2041–2058. Gauss, C. F., 1839, Allegemeine Theorie des Erdmagnetismus, in Gauss,C. F., and Weber,W., Resultate aus den Beobachtungen des magnetischen Vereins in Jahre 1838: Weidmann, Leipzig, p. 1–57. Gauss, C. F., and Weber, W., 1839, Resultate aus den Beobachtungen des magnetischen Vereins im Jahre 1838: Weidmann, Leipzig, 151 p. C71 unnumbered p. tables. Dieterich, Göttingen. Gauss, C. F., and Weber, W., 1840, Atlas des Erdmagnetismus: nachden elementen der theorie entworfen. Supplement zuden Resultaton aus den Beobachtungen des magnetischen. Vereins untermit wirkung von C. W. B. Goldschmidt: Wiedmann, Leipzig, 36 p. C8 unnumbered p. tables CXVIII plates. Gibson, M. O., 1941, Network adjustment by least squares- alternative formulation and solution by iteration: Geophysics, v. 6, no. 2, p. 168–179. Gilbert, W., 1600, De magnete, magnetisque corporibus, et de magneto magnete tellure; Physiologia nova, plurimis and argumentis, and experimentis demonstrata: Short, London, p. 240; reprinted Culture et Civilisation, Brussels (1967). Godwin, C. I., and Sinclair, A. J., 1979, Application of multiple regression analysis to drill-target selection, Casino porphyry copper-molybdenum deposit,Yukon Territory, Canada: Trans. Institution of Mining and Metallurgy, London, v. B88, no. 3, p. 93–106. Good, D. I., 1964, FORTRAN II trend-sur ...

bne: ... face program for the IBM 1620: Kansas Geol. Survey, Spec. Distrib. Publ. 14, 54 p. Govindaraju, K., 1987, 1987 compilation on Ailsa Craig granite AC-E with the participation of 128 GIT-IWG laboratories: Geostandards Newsletter, v. 11, no. 2, p. 203–255. Grant, F., 1957, A problem in the analysis of geophysical data: Geophysics, v. 22, no. 2, p. 309–344. Griffin, W. R., 1949, Residual gravity in theory and practice: Geophysics, v. 14, no. 1, p. 39–56. Griffiths, J. C., 1958, Petrography and porosity of the Cow Run Sand, St. Marys, West Virginia: Jour. Sedimentary Petrology, v. 28, no. 1, p. 15–30. Griffiths, J. C., 1967, Mathematical exploration strategy and decision-making: Proc. Panel Discussion 5, 7th World Petroleum Congress (Mexico City), p. 87–98. Grossman, M., 1971, Parametric curve fitting: Computer Jour. v. 14, no. 2, p. 169–172. Hald, A., 1998, A history of mathematical statistics from 1750 to 1930: John Wiley & Sons, New York, p. 795. Hall, A., 1969, Regional variation in the composition of British Caledonian Granites: Jour. Geology, v. 77, no. 4, p. 466–481. Halley, E., 1701, A new and correct chart shewing the variations of the compass in the western and southern oceans as observed in ye year 1700 by his Ma.ties Command by Edm. Halley: Mount and Page, London, 1 sheet. Hampel, F., 1971,Aqualitative definition of robustness: Ann. Math. Statistics, v. 42, no. 6, p. 1887–1896. Hampel, F. R., 1974, The influence curve and its role in robust estimation: Jour. Am. Stat. Assoc., v. 69, no. 346, p. 383–393. Harbaugh, J. W., 1963, BALGOL Program for trend-surface mapping using andIBM7090 computer: Kansas Geol. Survey, Spec. Distrib. Publ. 3, 54 p. Harbaugh, J.W., 1964, A computer method for four-variable trend analysis illustrated by a study of oil-gravity variations in southeastern Kansas: Kansas Geol. Survey Bull. 171, 58 p. Harbaugh, J. W., and Merriam, D. F., 1964, Trend surface analysis of regional and residual components of geologic structure in Kansas: Kansas Geol. Survey Spec. Dist. Publ. 11, 27 p. Harbaugh, J.W., and Preston, F.W., 1965, Fourier series analysis in geology, in Short Course and Symp. Computers and Computer Applications in Mining and Exploration, v. 1, College of Mines, Univ. Arizona, p. R1-R46. Harding, J. P., 1949, The use of probability paper for the graphical analysis of polymodal frequency distributions: Jour. Marine Biological Assoc. United Kingdom, v. 28, no. 1, p. 141–153. Hartley, H. O., 1961, The modified Gauss-Newton method for fitting nonlinear regression functions by least squares: Technometrics, v. 3, no. 2, p. 269–280. Haughton, D., 1997, Packages for estimating finite mixtures: a review: Am. Statistician, v. 51, no. 2, p. 194–205. Hellman, G., 1909, Magnetische Kartographie in historischkritischer Darstellung: Veröffentlichungen des Königlich Preußischen Meteorologischen Instituts no. 215, v. 3, no. 3, p. 1–61. Helsel, D. R., and Hirsch, R. M., 1992, Statistical methods in water resources: Elsevier, Amsterdam, 522 p. Henriques, A., 1958, The influence of cations on the optical properties of clinopyroxenes. II: Arkiv Mineralogi Geologi, v. 2, no. 4, p. 381–384. Hersh, A. H., 1934, Evolutionary relative growth in the Titanotheres: Am. Naturalist, v. 68, no. 719, p. 537–561. Hey, M. H., 1933, Studies on zeolites. Part V. Mesolite: Mineral. Magazine, v. 23, no. 144, p. 421–447. Hey, M. H., 1956, On the correlation of physical properties with chemical composition in multivariate systems: Mineral. Magazine, v. 31, no. 232, p. 69–95. Hinisch, M. J., and Talwar, P. P., 1975, A simple method for robust regression: Jour. Am. Stat. Assoc., v. 70, no. 349, p. 113–119. Hoaglin, D. C., and Welsch, R., 1978, The hat matrix in regression and ANOVA: Am. Statistician, v. 32, no. 1, p. 17–22. Hocking, R. R., 1983, Developments in linear regression methodology: 1959-1982 [with discussion]: Technometrics, v. 25, no. 3, p. 219–249. Hocking, R. R., and Leslie, R. N., 1967, Selection of the best subset in regression analysis: Technometrics, v. 9, no. 4, p. 531–540. Hoerl, A. E., 1962, Application of ridge analysis to regression problems: Chemical Engineering Progress, v. 58, p. 54–59. Hoerl, A. E., and Kennard, R.W., 1970a, Ridge regression: biased estimation for nonorthogonal problems: Technometrics, v. 12, no. 1, p. 55–67. Hoerl, A. E., and Kennard, R.W., 1970b, Ridge regression: applications to nonorthogonal problems: Technometrics, v. 12, no. 1, 69–82. Hood, K., Nix, B. A. J., and Iles, T. C., 1999, Asymptotic information and variance-covariance matrices for the linear structural model: The Statistician, v. 48, no. 4, p. 477–493. Horner, J. K., 1836, Tellurischer Magnetismus, in Gehler, J. S. T., ed., Physikalisches Wörterbuch oder Versuch einer Erklärung der vornehmsten Begriffe und Kunstwörter der Naturlehre, v. 6, abt. 2, taf XXIV. Schwickert, Leipzig. Howarth, R. J., 1984, Statistical applications in geochemical prospecting: a survey of recent developments: Jour. Geochem. Exploration, v. 21, no. 1-3, p. 41–61. Howarth, R. J., in press a, Fitting geomagnetic fields before the invention of least-squares: I. Henry Bond's predictions (1636- 1668) of the change in magnetic declination in London, Ann. Science. Howarth, R. J., in press b, Fitting geomagnetic fields before the invention of least-squares: II.WilliamWhiston's isoclinic maps of southern England (1719 and 1721): Ann. Science. Howarth, R. J., and Koch, G. S., Jr., 1986, Problems of using rock volume data in predictive resource studies: Econ. Geology, v. 81, no. 3, p. 617–626. Howarth, R. J., and McArthur, J. M., 1997, Statistics for strontium isotope stratigraphy: a robust LOWESS fit to the marine Srisotope curve for 0 to 206 Ma, with look-up table for derivation of numeric age: Jour. Geology, v. 105, no. 4, p. 441–456. Huber, P. J., 1964, Robust estimation of a location parameter: Ann. Math. Statistics, v. 35, no. 1, p. 73–101. Huber, P. J., 1973, Robust regression: asymptotics, conjectures and Monte Carlo: Ann. Statistics, v. 1, no. 5, p. 799–821. Humboldt, F. H. A. von, and Biot, J. B., 1804, Sur les variations du magnétisme terrestre, àdifférentes latitudes: Jour. Physique, v. 59, p. 429–450. Imbrie, J., and Kipp, N. G., 1971, A new micropalaeontological method for quantitative paleoclimatology: application to a Late Pleistocene Caribbean core, in Turekian, K., ed., The Late Cenozoic Glacial Ages: Yale Univ. Press, New Haven, Connecticut, p. 71–181. Inman, J. R., 1975, Resistivity inversion with ridge regression: Geophysics, v. 40, no. 5, p. 798–817. Ivory, J., 1825, On the method of least squares: Tilloch's Philosophical Jour. v. 65, p. 3–10, 81-86, 161-168. Jaeckel, L. A., 1972, Estimating regression coefficients by minimizing the dispersion of the residuals: Ann. Math. Statistics, v. 43, p. 1449–1458. James, W. R., 1966, FORTRAN IV program using double Fourier Series for surface fitting of irregularly spaced data: Kansas Geol. Survey, Computer Contrib. 5, 19 p. Jones, H. E., 1937, Some geometrical considerations in the general theory of fitting lines and planes: Metron, v. 13, no. 1, p. 21–30. Jones, T. A., 1972, Multiple regression with correlated independent variables: Math. Geology, v. 4, no. 3, p. 203–218. Jones, T. A., 1979, Fitting straight lines when both variables are subject to error. I. Maximum liklihood and least squares estimation: Math. Geology, v. 11, no. 1, p. 1–25. Jorgenson, D. W., 1961, Multiple regression analysis of a Poisson process: Jour. Am. Stat. Assoc., v. 56, no. 294, p. 235–245. Journel, A. G., and Huijbregts, C. J., 1978, Mining geostatistics: Academic Press, London, 600 p. Jupp, D. L., 1976, B-Splines for smoothing and differentiating data sequences: Math. Geology, v. 8, no. 3, p. 243–266. Jupp, D. L., and Stewart, I. C. F., 1974, A piecewise exponential model for seismic well-logging data: Math. Geology, v. 6, no. 1, p. 33–45. Juran, J. M., 1991,WorldWar II and the quality movement: Quality Progress, v. 24, no. 12, p. 19–24. Jurecková, J., 1971, Nonparametric estimate of regression coeffi-cients: Ann. Math. Statistics, v. 42, no. 4, p. 1328–1338. Kawasaki, T., and Matsui, Y., 1978, Nonlinear least squares technique in element prediction equilibria: Geochemical Jour., v. 12 no. 3, p. 173–181. Kendall, M. G., 1946, The advanced theory of statistics: Griffin, London, v. 2, 521 p. Kent, J. T., Watson, G. S., and Onstott, T. C., 1990, Fitting straight lines and planes with an application to radiometric dating: Earth and Planetary Science Letters, v. 97, no. 1, p. 1–17. Kimeldorf, G. S., and Wahba, G., 1970, A correspondence between Bayesian estimation on stochastic processes and smoothing by splines: Ann. Math. Statistics, v. 41, no. 2, p. 495–502. Knuth, D. E., and Pardo, L. T., 1980, The early development of programming languages, in Metropolis, N., Howlett, J., and G.-C., Rota, eds., A History of Computing in the Twentieth Century: Academic Press, New York, p. 197–273. Koch, G. S., Jr., and Link, R. F., 1967, Geometry of metal distribution in five veins of the Fresnillo Mine, Zacatecas, Mexico: Bur. Mines Rept. Invest. no. 6919, U.S. Depart. Interior, Bur. Mines, Washington, p. 1–64. Koch, G. S., Jr., and Link, R. F., 1970-71, Statistical analysis of geological data: 2 vols. John Wiley & Sons, New York, 375 p., 438 p. Koch, G. S., Jr., Link, R. F., and Schuenemeyer, J. H., 1972, Computer programs for geology: Artronic Information Systems, New York, 142 p. Kohlrausch, F., 1873, An introduction to physical measurements with appendices on absolute electrical measurement, etc.: Waller, T. H., and Procter, H. R., transl., Churchill, London, 249 p. Kottegoda, N. T., and Rosso, R., 1998, Statistics, probability, and reliability for civil and environmental engineers: McGraw-Hill Book Co., New York, 735 p. Krafft, W. L., 1809, Essai sur une loi hypothétique des incolnations de l'aguille aimantée en differents endroits de la terre: Mémoires de l'Académic Impériale des Sciences de St. Pétersbourg, v. l, p. 248–270. Krige, D. G., 1951, A statistical approach to some basic mine valuation problems on the Witwatersrand: Jour. Chem. Metallurgical and Mining Soc. South Africa, v. 52, p. 119–139. Krige, D.G., 1960, On the departure of ore value distributions from the lognormal model in South African Gold Mines: Jour. South African Inst. Mining and Metallurgy, v. 64, no. 4, p. 231–244. Krige, D. G., 1966, Two-dimensional weighted moving average trend surfaces for ore valuation [with discussion]: Jour. South African Inst. Mining and Metallurgy. Spec. Symp. issue, p. 13–79, p. 379-380. Krige, D. G., and Ueckermann, H. J., 1963, Value contours and improved regression techniques for ore reserve valuations: Jour. South African Inst. of Mining and Metallurgy, v. 63, no. 10, p. 429–452. Krumbein, W. C., 1937, Sediments and exponential curves: Jour. Geology, v. 45, no. 6, p. 577–601. Krumbein, W.C., 1952, Principles of faciesmapinterpretation: Jour. Sedimentary Petrology, v. 22, no. 4, p. 200–211. Krumbein, W. C., 1956, Regional and local components in facies maps: Am. Assoc. Petroleum Geologists Bull. v. 40, no. 9, p. 2163–2194. Krumbein, W. C., 1959a, Trend surface analysis of contour-type maps with irregular control-point spacing: Jour. Geophysical Research, v. 64, no. 7, p. 823–834. Krumbein, W. C., 1959b, The sorting out of geological variables illustrated by regression analysis of factors controlling beach firmness: Jour. Sedimentary Petrology, v. 29, no. 4, p. 575–587. Krumbein, W. C., and Pettijohn, F. J., 1938, Manual of sedimentary petrography: Appleton-Century, New York, 549 p. Krumbein, W. C., and Sloss, L. L., 1951, Stratigraphy and sedimentation: W. C. Freeman & Co., San Francisco, 497 p. Krumbein, W. C., and Sloss, L. L., 1958, High-speed digital computers in stratigraphic and facies analysis: Am. Assoc. Petroleum Geologists Bull., v. 42, no. 11, p. 2650–2669. Lalande, J. J. LeF, 1771-81, Astronomie (2nd edn.): 4 vols. Desaint, Paris, 608 p., 803 p., 840 p., 788 p. Lambert, J. H., 1765, Theorie der Zuverlässigkeit der beobachtungen und Versuche, in Beyträge zum Gebrauche der Mathematik und deren Anwendung, v. 1: Verlage des Buchlagens der Realschule, Berlin, p. 424–488. Laplace, P. S. de, 1786,Mémoire sur la figure de la terre:Mémoires de l'Académie royale des Sciences de Paris, v. for 1783, p. 17–46. Laplace, P. S. de, 1788, Théorie de Jupiter et de Saturne:Mémoires de l'Académie royale des Sciences de Paris, v. for 1785, p. 33–160. Laplace, P. S. de, 1793, Sur quelques points du Systéme du monde: Mémoires de l'Académie royale des Sciences de Paris, v. for 1789, p. 477–558. Laplace, P. S. de, 1799, Traité de Méchanique Céleste: v. 2, Crapelet, Paris, 382 p. Laplace, P. S. de, 1819, Application du calcul des probabilités aux opérations géodésiques de la m´eridienne de France: Bulletin des Sciences de la Société Philomathique de Paris, v. for 1819, p. 137–139. Laslett, G. M., 1994, Kriging and splines: An empirical comparison of their predictive performance in some applications [with discussion]: Jour. Am. Stat. Assoc., v. 89, no. 3, p. 391–409. Le Maitre, R. W., 1982, Numerical petrology. Statistical interpretation of geochemical data: Developments in Petrology 8, Elsevier, Amsterdam, 281 p. Legendre, A. M., 1805, Nouvelles méthodes pour la determination des orbites des cométes: Courcier, Paris, 80 p. Legge, J. A., Jr., and Rupnik, J. J., 1943, Least squares determination of the velocity function V D V0 C kz for any set of time depth data: Geophysics, v. 8, no. 4, p. 356–362. Lenth, R. V., 1977, Robust splines: Communications in Statistics, Theory and Methods, v. A6, no. 9, p. 847–854. Lepeltier, C., 1969,Asimplified statistical treatment of geochemical data by graphical representation: Econ. Geology, v. 64, no. 5, p. 538–550. Li, Y., 1985, Recognition of mineralized belts and lithologic patterns in the Silver City-South Mountain region, Idaho, in terms of geochemical reconnaissance data: Math. Geology, v. 17, no. 3, p. 243–265. Lines, L. R., and Treitel, S., 1985, A review of nonlinear regression and its applications to geophysical inverse problems, in Anderson, O.D., Ord, J. K., and Robinson, J. E., eds., Time series analysis: theory and practice 6. Hydrological, geophysical and spatial applications: Elsevier, Amsterdam, p. 133–178. Link, R. F., and Koch, G. S., Jr., 1962, Quantitative areal modal analysis of granitic complexes: discussion: Geol. Soc. America Bull. v. 73, no. 3, p. 411–414. Link, R. F., and Koch, G. S., Jr., 1974, A comparison of three methods of smoothing for ore estimation, in Johnson, T. B., and Gentry, D. W., eds. Proc. 12th Intern. Symp. Applications of Computers and Mathematics in the Minerals Industry, v. 2: Colorado Sch. Mines, Golden, Colorado, p. F95–F107. Link, R. F., Koch, G. S., Jr., and Gladfelter, G.W., 1964, Computer methods of fitting surfaces to assay and other data by regression analysis: Bur. Mines Rept. Invest. no. 6508, U.S. Depart. Interior, Bur. Mines, Washington, 69 p. Link, R. F., Yabe, N. N., and Koch, G. S., Jr., 1966, A computer method of fitting surfaces to assay and other data in three dimensions by quadratic regression analysis: Bur. Mines Rept. Invest. no. 6876,U.S. Depart. Interior, Bur. Mines, Washington, 42 p. Lister, B., 1982, Evaluation of analytical data: a practical guide for geoanalysts: Geostandards Newsletter, v. 6, no. 2, p. 175–205. Lister, B., 1984, A note on robust estimators: Geostandards Newsletter, v. 8, p. 171–172. Lloyd, H., Sabine, E., and Ross, J. C., 1836, Observations on the direction and intensity of the terrestrial magnetic force in Ireland, in Rept. 5th meeting of the British Association for the Advancement of Science, Dublin, 1835: Murray, London, p. 117–162. Longley, J.W., 1967, An appraisal of least squares programs for the electronic computer from the point of view of the user: Jour. Am. Stat. Assoc. v. 62, no. 319, p. 819–841. Ludwig, K. R., 1980, Calculations of uncertainties of U-Pb isotope data: Earth and Planetary Science Letters, v. 46, no. 2, p. 212–220. Ludwig, K. R., 1993, ISOPLOT:Aplotting and regression program for radiogenic isotope data.:U.S. Geol. SurveyOpen-File Rept. 91-445, 41 p. Ludwig, K. R., and Titterington, D. M., 1994, Calculation of 230Th/U isochrons, ages, and errors: Geochim. et Cosmochem. Acta, v. 58, no. 22, p. 5031–5042. Ludwig, K. R., Halley, R. B., Simmons, K. R., and Peterman, Z. E., 1988, Strontium-isotope stratigraphy of Enewetak Atoll: Geology, v. 16, no. 2, p. 173–177.

bne: Mahon, K. I., 1996, The New York regression: application of an improved statistical method to geochemistry: Intern. Geology Review, v. 38, no. 4, p. 293–303. Maire, C. and Boscovich, R. J., (1755) 1770, Voyage astronomique et geographique, dans l' йtat de l'eglise, entrepris par l'ordre et sous les auspices du Pape Benoit XIV, Pour mesurer deux dйgrйs du mйridien, et corriger la Carte de l'Etat ecclйsiastique, Par les PP. Maire et Boscovich de la Companie de Jesus, Traduit du latin, Augmentй de notes et d'extraits de nouvelles mesures de dйgrйs faites en Italie, en Allemagne, en Hongrie et en Amйrique. Avec une nouvelle Carte des Etats du Pape levйe gйometriquement: Tillard, Paris, 512 p. Mandelbaum, H., 1963, Statistical and geological implications of trend mapping with nonorthogonal polynomials: Jour. Geophysical Research, v. 68, no. 2, p. 505–519. Mann, C. J., 1987, Misuses of linear regression in the earth sciences, in Size, W. B., ed., Use and Abuse of Statistical Methods in the Earth Sciences, Intern. Assoc. Math. Geology Studies in Mathematical Geology 1: Oxford Univ. Press, Oxford, p. 74–106. Marquardt, D. W., 1963, An algorithm for the estimation of nonlinear parameters: Jour. Soc. Industrial and Applied Mathematics, v. 11, no. 2, p. 431–441. Marquardt, D.W., 1970, Generalized inverses, ridge regression, biased linear estimation, and nonlinear estimation: Technometrics, v. 12, no. 4, p. 591–612. Matheron, G., 1962-3, Traitй de gйostatistique appliquйe: 2 vols. Technip, Paris, 334 p., 172 p. Matheron, G., 1963, Principles of geostatistics: Econ. Geology, v. 58, p. 1246–1266. Matheron, G., 1965, Les variables rйgionalisйes et leur estimation: Masson, Paris, 212 p. Matheron, G., 1981, Splines and Kriging: their formal equivalence, in Merriam, D. F., ed., Down-to-Earth Statistics. Solutions Looking for Geological Problems: Syracuse Univ. Geology Contribution 8, p. 77–95. Maupertuis, P. L.M. de, 1740, La figure de la Terre dйterminйe par MM.de l'Acadйmie R. des Sciences qui ont mesurй le degrйdu Meridien au Cercle polaire: Histoire de l'Acadйmie royale des inscriptions et belles lettres, avec les Mйmoires de litterature tirйes des registres de cette acadйmie, v. for 1738, p. 90–96, p. 389-466. Mayer, T., 1750, Abhandlungьber die Umwalzung des Monds um seine Axe und die scheinbare Bewegung der Mondsflecten: Kosmographische Nachrichten und Sammlungen, v. for 1748, p. 52–183. McArthur, J. M., Howarth, R. J., and Bailey, T. R., 2001, Strontium isotope stratigraphy: LOWESS Version 3. Best-fit to the marine Sr-isotope curve for 0 to 509 Ma and accompanying lookup table for deriving numerical age: Jour. Geology, v. 109, no. 2, p. 155–170. McCammon, R. B., 1969, FORTRAN IV program for nonlinear estimation: Kansas Geol. Survey, Computer Contr. 34, 20 p. McCammon, R. B., 1973, Nonlinear regression for dependent variables: Math. Geology, v. 5, no. 4, p. 365–375. McIntyre, D. B., 1963, Program for computation of trend surfaces and residuals of degree 1 through 8: Tech. Rept. 4, Depart. Geology, Pomona College, Claremont, California, 24 p. McIntyre, G. A., Brooks, C., Compston, W., and Turek, A., 1966, The statistical assessment of Rb-Sr isochrons: Geophys. Research, v. 71, no. 22, p. 5459–5468. McKenzie, D.P., 1987, Edward Crisp Bullard: biographical memoirs of Fellows of the Royal Society, v. 33, p. 67–98. McLaughlin, O. M., McArthur, J. M., Thirlwall, M. F., Howarth, R. J., Burnett, J., Gale, A. S., and Kennedy, W. J., 1995, Sr isotope evolution of Maastrichtian seawater determined from the chalk of Hemmoor, Germany: Terra Nova, v. 7, no. 5, p. 491–499. Meckel, L. D., 1967, Origin of Pottsville conglomerates (Pennsylvanian) in the central Appalachians: Geol. Soc.America Bull., v. 78, no. 2, p. 223–258. Mendoza, C. E., 1986, Smoothing unit vector fields: Math. Geology, v. 18, no. 3, p. 307–322. Merriam, D. F., 1981, Roots of quantitative geology, in Merriam, D. F., ed., Down-to-Earth Statistics. Solutions Looking for Geological Problems: Syracuse Univ. Geology Contr. 8, p. 1–15. Merriam, D. F., and Doria-Medina J. H., 1968, Analisis de tendencias polinomias y de Fourier aplicados a la informacion estratigrafica: Boletin del Inst. Boliviano del Petroleo, v. 8, no. 1, p. 59–74. Merriam, D. F., and Sneath, P. H. A., 1966, Quantitative comparison of contour maps: Jour. Geophysical Research, v. 71, no. 4, p. 1105–1115. Merriman, M., 1877,Alist of writings relating to the method of least squares, with historical and critical notes: Trans. Connecticut Academy of Arts and Sciences, v. 4, no. for 1877-82, pt. 1, p. 151–232. Mickey, M. R., Jr., and Jesperesen, H. W., 1954, Some statistical problems of uranium exploration. Final Tech. Rept: U.S. Atomic Energy Commission, Rept. RME-3105 (Oak Ridge, Tennessee), 78 p. Miesch, A. T., and Connor, J. J., 1967, Stepwise regression in trend analysis: Kansas Geol. Survey, Computer Contr. 12, p. 16–18. Miesch, A. T., and Connor, J. J., 1968, Stepwise regression and nonpolynomial models in trend analysis: Kansas Geol. Survey, Computer Contr. 27, 40 p. Miller, R. L., 1956, Trend surfaces: their application to analysis and description of environments of sedimentation. I. The relation of sediment-size parameters to current-wave systems and physiography: Jour. Geology, v. 64, no. 5, p. 425–446. Miller, R. L., and Kahn, J. S., 1962, Statistical analysis in the geological sciences: John Wiley & Sons, New York, 483 p. Montgomery, D. C., Peck, E. A., and Simpson, J. R., 1998, Multicollinearity and biased estimation in regression, inWadsworth, H. M., Jr. ed., Handbook of Statistical Methods for Engineers and Scientists (2nd edn.): McGraw-Hill Book Co., New York, p. 16.3–16.27. Mosteller, F., 1947, On some useful inefficient statistics: Ann. Math. Statistics, v. 17, no. 4, p. 377–408. Mьller, H.-G., 1987,Weighted local regression and kernel methods for nonparametric curve fitting: Jour. Am. Stat. Assoc. v. 82, no. 2, p. 231–238. Murtagh, F., 1990, Linear regression with errors in both variables: a short review, in Jascheck, C., and Murtagh, F., eds., Errors, bias and uncertainties in astronomy: Cambridge Univ. Press, Cambridge, p. 385–391. Murthy, V. R., and Compston, W., 1965, Rb-Sr ages of chondrules and carbonaceous chondrites: Jour. Geophysical Research, v. 70, no. 20, p. 5297–5307. Nackowski, M. P., Mardirosian, C. A., and Botbol, J. M., 1967, Trend surface analysis of trace chemical data, Park City District, Utah: Econ. Geology, v. 62, no. 8, p. 1072–1087. Newton, I., 1726 (1934), Book III. System of theWorld (in mathematical treatment), in Motte, A. transl., Cajori, F., ed., Sir Isaac Newton's Mathematical Principles of Natural Philosophy and his System of theWorld, v. II: reprinted Univ. California Press, Berkeley, Los Angeles, p. 397–680. Nichol, I., and Webb, J. S., 1967, The application of computerized mathematical and statistical procedures to the interpretation of geochemical data [with discussion]: Proc. Geol. Society, London, no. 1642, p. 186–199. Nichol, I., Garrett, R. G., and Webb, J. S., 1969, The role of some statistical and mathematical methods in the interpretation of regional geochemical data: Econ. Geology, v. 64, no. 2, p. 204–220. Norcliffe, G. B., 1969, On the use and limitations of trend surface analysis: Can. Geographer, v. 13, no. 4, p. 338–348. O'Leary, M., Lippert, R. H., and Spitz, O. T., 1966, FORTRAN IV and MAP program for computation and plotting of trend surfaces for degrees 1 through 6: Kansas Geol. Survey, Computer Contr. 3, 47 p. O'Neill, M., Sinclair, I. G., and Smith, F. J., 1969, Polynomial curve fitting when abscissas and ordinates are both subject to error: Computer Jour., v. 12, no. 1, p. 53–57. Oldham, C. H.G., and Sutherland, D. B., 1955, Orthogonal polynomials: their use in estimating the regional effect: Geophysics, v. 20, no. 2, p. 295–306. Olea, R. A., 1972, Application of regionalized variable theory to automatic contouring: Spec. Rept.Am.Petroleum Inst. Research Project no. 131, Univ. Kansas Center for Research Inc. and Kansas Geol. Survey, Lawrence, Kansas, 191 p. Olmstead, P. S., 1967, Special Memorial Issue: Industrial Quality Control, v. 24, no. 2, p. 72–122. Omori, F., 1894, On after-shocks: Seismological Jour. Japan, v. 3, p. 71–80. Ondrick, C. W., and Griffiths, J. C., 1969, Trend surface analysis applied to the Rensselaer graywacke and its implications to the Taconics: Jour. Sedimentary Petrology, v. 39, no. 1, p. 176–186. Pan, G., and Harris, P. D., 1991, Geology-exploration endowment models for simultaneous estimation of discoverable mineral resources and endowment: Math. Geology, v. 23, no. 4, p. 507–540. Pearson, K., 1893, Assymetrical frequency curves: Nature, v. 48, no. 1252, p. 615–616. Pearson, K., 1894, Contribution to the mathematical theory of evolution: Philosophical Trans. Royal Society, London, v. A185, no. for 1893, p. 71–110. Pearson, K., 1896, Contributions to the mathematical theory of evolution. II. Skew variation in homogenious material: Royal Society, London, Philosophical Trans., v. A186, no. for 1895, p. 343–414. Pearson, K., 1901, On lines and planes of closest fit to systems of points in space: Philosophical Magazine, Ser. 6, v. 2, no. 11, p. 559–572. Pearson, K., 1920, Notes of the history of correlation: Biometrika, v. 13, p. 25–45. Pell, J., 1673-76, Collections relating to commissions to enquire into the methods for finding the longitude: British Library Add. ms. 4393, London, ff. 26–106. Pelletier, B. C., 1958, Pocono paleocurrents in Pennsylvania and Maryland: Geol. Soc. America Bull., v. 69, no. 8, p. 1033–1064. Perry, K., Jr., 1967, Methods of petrologic calculation and the relationship between mineral and bulk chemical composition: Univ. Wyoming Contr. to Geology, v. 6, no. 1, p. 5–38. Perry, K., Jr., 1968, A computer program for representation of mineral chemical analyses in terms of end members: Univ. Wyoming Contr. to Geology, v. 7, no. 1, p. 7–14. Potter, P. E., 1955,The petrology and origin of the Lafayette Gravel: Part I, mineralogy and petrology: Jour. Geology, v. 63, no. 1, p. 1–38. Priestley, M. B., and Chao, M. T., 1972, Nonparametric function fitting: Jour. Royal Stat. Society, v. B34, no. 3, p. 384–392. Puissant, L., 1819, Traitй de gйodйsie, ou exposition des mйthodes trigonomйtriques et astronomiques, applicables аla construction du canevas des cartes topographiques: 2 vols. Courcier, Paris, 426 p., 386 p. Puissant, L., 1842, Trait йde gйodйsie, ou exposition des mйthodes trigonomйtriques et astronomiques, applicables аla construction du canevas des cartes topographiques (2nd edn.) 2 vols. Courcier, Paris, 515 p., 496 p. Quetelet, A., 1841, Second Mйmoire sur le Magnйtisme Terrestre en Italie: Nouveaux Mйmoires de l'Academie royale des Sciences et des Belles-Lettres de Bruxelles, v. 13, p. 1–27. Ralston, M. L., and Jennrich, R., 1978, Derivative-free nonlinear regression, in Hogben, D., ed., Proc. Computer-Science and Statistics.Tenth Ann. Symp. Interface:U.S. Government Printing Office, Washington, D.C., p. 312–322. Rao, C. R., 1948, The utilization of multiple measurements in problems of biological classification: Jour. Royal Stat. Society, v. B10, no. 2, p. 159–203. Ray, R. D., 1985, Correlation of systematic error in magnetic surveys, an application of ridge regression and sparse matrix theory: Geophysics, v. 50, no. 11, p. 1721–1731. Read, L. J., and Berkson, J., 1929, The application of the logistic function to experimental data: Jour. Physical Chemistry, v. 33, p. 760–779. Read, W. A., and Merriam, D. F., 1966, Trend surface analysis of stratigraphic thickness data from some Namurian rocks east of Stirling, Scotland: Scottish Jour. Geology, v. 2, no. 1, p. 96–100. Renner, R. M., 1991, An examination of the use of the logratio transformation for the testing of endmember hypotheses: Math. Geology, v. 23, no. 4, p. 549–563. Renner, R. M., 1993, A constrained least-squares subroutine for adjusting negative estimated element concentrations to zero: Computers & Geosciences, v. 19, no. 9, p. 1351–1360. Renner, R. M., Glasby, G. P., and Szefer, P., 1998, Endmember analysis of heavy-metal pollution in surficial sediments from the Gulf of Gdansk and the southern Baltic Sea off Poland: Applied Geochemistry, v. 13, no. 3, p. 313–318. Renner, R. M., Glasby, G. P., and Walter, P., 1997, Endmember analysis of metalliferous sediments from the Galapagos Rift and East Pacific Rise between 2є N. and 42є S: Applied Geochemistry, v. 12, no. 3, p. 383–395. Richardson, W. A., 1923, The frequency-distribution of igneous rocks. Part II. The laws of distribution in relation to petrogenetic theories: Mineral. Magazine, v. 20, no. 100, p. 1–19. Ripley, B.D., 1981, Spatial statistics: JohnWiley & Sons, NewYork, 252 p. Ripley, B. D., and Thompson, M., 1987, Regression techniques for the detection of analytical bias: The Analyst, London, v. 112, p. 377–383. Ritter, K., 1969, Generalized spline interpolation and non-linear Programming, in Schoenberg, I. J., ed., Approximation with Special Emphasis on Spline Functions: Academic Press, New York, p. 75–118. Riu, J., and Rius, F. X., 1995, Univariate regression models with errors in both axes: Jour. Chemometrics, v. 9, p. 343–362. Rock, N. M. S., and Duffy, T. R., 1986, REGRES:AFORTRAN-77 program to calculate nonparametric and structural parametric solutions to bivariate regression equations: Computers & Geosciences, v. 12, no. 6, p. 807–818. Roeder, K., 1994, A graphical technique for determining the number of components in a mixture of normals: Jour. Am. Stat. Assoc., v. 89, no. 2, p. 487–495. Rollinson, H. R., 1993, Using geochemical data: evaluation, presentation, interpretation: Longman, Harlow, 352 p. Roman, I., 1933, Least squares in practical geophysics, in Geophysical Prospecting 1932: Am. Inst. of Mining Engineers, New York, p. 460–506. Rose, A. W., Dahlberg, E. C., and Keith, M. L., 1970, A multiple regression technique for adjusting background values in stream-sediment geochemistry: Econ. Geology, v. 65, no. 2, p. 156–165. Rousseeuw, P. J., 1983, Regression techniques with a high breakdown point: Inst. Mathematics and Statistics Bull., v. 12, p. 155. Rousseeuw, P. J., 1984, Least median of squares regression: Jour. Am. Stat. Assoc., v. 79, no. 388, p. 871–880. Ryan, T. A., Jr., 1975, Robust regression-bounded leverage: Proc. Stat. Computing Section (Alexandria, Virginia), Am. Stat. Assoc., p. 138–141. Sabine, E., 1837, Observations on the direction and intensity of the terrestrial magnetic force in Scotland, in Rept. 6th Meeting of the British Association for the Advancement of Science, (Bristol, 1836): Murray, London, p. 97–119. Sabine, E., 1839a, Report on the variations of the magnetic intensity observed at different points of the Earth's surface, in Rept. 8th meeting of the British Association for the Advancement of Science (Newcastle-on-Tyne, 1838): Murray, London, p. 1–85, p. 497-500. Sabine, E., 1839b, A memoir on the magnetic isoclinal and isodynamic lines in the British Islands, from observations by Professors Humphry Lloyd and John Phillips, RobertWere Fox, Esq., Captain James Clark Ross, R. N., and Major Edward Sabine, R. A., in Rept. 8th meeting of the British Association for the Advancement of Science. (Newcastle-on-Tyne, 1838): Murray, London, p. 49–196, 318-320. Santos Oliveira, J. M., 1979, Trend-surface analysis in geochemical propsecting data, Arouca-Castro Daire region, northern Portugal: Chem. Geology, v. 24, no. 3-4, p. 271–291. Schlee, J., 1957, Upland gravels of southern Maryland: Geol. Soc. America Bull., v. 68, no. 10, p. 1371–1410. Schmid, K., 1934, Biometrische Untersuchungen an Foraminiferen aus dem Phacen von Ceram: Eclogae Geologicae Helvetiae, v. 27, no. 1, p. 46–128. Schoenberg, I. J., 1946, Contributions to the problem of approximation of equidistant data by analytic functions: Quart. Applied Mathematics, v. 4, no. 1, p. 45–99, 112-141. Shewhart, W. A., 1933, Economic control of manufactured product: Van Nostrand, New York, 356 p. Sheynin, O. B., 1993, On the history of the principle of least squares: Archive for History of Exact Sciences, v. 46, p. 39–54. Simpson, S. M., Jr., 1954, Least squares polynomial fitting to gravitational data and density plotting by digital computers: Geophysics, v. 19, no. 2-3, p. 255–269, 644. Sinclair, A. J., and Woodsworth, G. J., 1970, Multiple regression as a method of estimating exploration potential in an area near Terreace, B.C.: Econ. Geology, v. 65, no. 8, p. 998–1003. Smalley, P. C., Higgins, A. C., Howarth, R. J., Nicholson, H., Jones, C. E., Swinburne, N. H. M., and Bessa, J., 1994, Marine Strontium isotopes: a Phanerozoic seawater curve for practical sediment dating and correlation: Geology, v. 22, no. 5, p. 431–434. Smith, J. R., 1986, From plane to spheroid. Determining the Figure of the Earth from 3000 B.C. to the 18th Century Lapland and Peruvian Survey Expeditions: Landmark Enterprises, Rancho Cordova, California, 219 p. Smith, J. V., Stevenson, D. A., Howie, R. A., and Hey, M. H., 1969, Relations between cell dimensions, chemical composition, and site preference of orthopyroxene: Mineral. Magazine, v. 37, no. 285, p. 90–114. Snee, R.D., 1983, Discussion:Technometrics, v. 26, no. 3, p. 230–237. Sobel, D., 1996, Longitude: Fourth Estate, London, 184 p. Somerville, M., 1831, Mechanisms of the Heavens: Murray, London, 1xx C 621 p. Stigler, S. M., 1973, Simon Newcomb, Percy Daniell, and the history of robust estimation 1885-1920: Jour. Am. Stat. Assoc., v. 86, no. 4, p. 872–879. Stigler, S. M., 1986, The history of statistics. The measurement of uncertainty before 1900: Belknap Press, Cambridge, Massachusetts, 410 p. Sytarno, D., and Vozoff, K., 1989, Robust M-estimation of magnetotelluric impedance tensors: Exploration Geophysics, v. 20, no. 3, p. 383–398. Tanner, W. F., Evans, R. G., and Holmes, C. W., 1963, Low-energy coast near Romano, Florida: Jour. Sedimentary Petrology, v. 33, no. 3, p. 713–722. Tap, J., and Bond, H., 1648, The Sea-Mans Kalendar or, An Ephemerides of the Sunne, moone, and certaine of the most notable fixed Starres.As also aTable of the Longitude and Latitude, of all the most eminent Places of theWorld, very exactly calculated by John Tap. The Twentieth Impression, newly Corrected and Inlarged with many additions ::: Also a Discovery of finding the long hidden Secret of longitude by Henry Bond, Teacher of the Mathematickes at the Bulwark-gate neere the Tower: Hurlock, London, 180 p. Taylor, E. G. R., 1954, The mathematical practitioners of Tudor and Stuart England: Cambridge Univ. Press for The Institute of Navigation, Cambridge, 443 p. Tilling, L., 1975, Early experimental graphs: British Jour. History of Science, v. 8, no. 30, p. 193–213. Tinkler, K. J., 1969, Trend surfaces with low explanations; the assessment of their significance:Am.Jour. Science, v. 267, no. 1, p. 114–123. Titterington, D. M., and Halliday, A. N., 1979, On the fitting of parallel isochrons and the method of maximum likelihood: Chem. Geology, v. 26, no. 3-4, p. 183–195. Titterington, D. M., Smith, A. F. M., and Makov, U. E., 1986, Statistical analysis of finite mixture distributions: John Wiley & Sons, Chichester, 237 p. Todhunter, I., 1873, A history of the mathematical theories of attraction and the Figure of the Earth: 2 vols. Macmillan, London, 476 p., 508 p.; reprinted, Dover, New York (1962). Troutman, B. M., and Williams, G. P., 1987, Fitting straight lines in the earth sciences, in Size, W. B., ed., Use and Abuse of Statistical Methods in the earth Sciences: Intern. Assoc. Math. Geology Studies in Mathematical Geology, 1: Oxford Univ. Press, Oxford, p. 107–128. Tukey, J. W., 1960, A survey of sampling from contaminated distributions, in Olkin, I., ed., Contributions to Probability and Statistics: Stanford Univ. Press, Stanford, California, p. 445–485. Tukey, J.W., 1972, Discussion of Mathematical Problems in Geology by F. P. Ageterberg and S. C. Robinson: Intern. Stat. Inst. Bull., v. 44, Proc. of 38th Sess. 1971, Book 1, p. 596. Tukey, J. W., (1970) 1977, Exploratory data analysis: Addison-Wesley, Reading, Massachusetts, 506 p.; early version of text privately circulated in 1970. Unwin, D. J., 1970, Percentage reduction in sums of squares in trend surface analysis: Area, v. 2, no. 1, p. 25–28. Van Orstrand, C. E., 1926, Some evidence on the variation of temperature with geologic structure in California and Wyoming oil districts: Econ. Geology, v. 21, no. 2, p. 145–165. Velleman, P. F., and Hoaglin, D. C., 1981, Applications, basics, and computing of exploratory data analysis: Duxbury Press, Boston, Massachusetts, 354 p. Vincent, P., 1986, Differentiation of modern beach and coastal dune sands; a logistic regression approach using the parameters of the hyperbolic function: Sedimentary Geology, v. 49, no. 3-4, p. 167–176. Vinod, H. D., and Ullah, A., 1981, Recent advances in regression methods: Dekker, New York, 361 p. Vistelius, A. B., 1967, Mathematical techniques in making geological interpretation: Proc. Panel Discussion 5, 7th World Petroleum Congress (Mexico City), p. 13–24. Vistelius, A. B., and Romanova, M. A., 1964, Distribution of the heavy fraction in sands from deposits of the central Kara Kum [in Russian]: Doklady Akademii Nauk SSSR, v. 158, no. 4, p. 860–864. Waddington, C. H., 1929, Notes on graphical methods of recording the dimensions of ammonites: Geol. Magazine, v. 66, no. 778, p. 180–186. Wahba, G., 1975, Smoothing noisy data by spline functions: Numerische Mathematik, v. 24, p. 383–393. Wald, A., 1940, The fitting of straight lines if both variables are subject to error: Ann. Math. Statistics, v. 11, no. 3, p. 284–300. Welch, A. H., Lico, M. S., and Hughes, J. L., 1988, Arsenic in ground water of the western United States: GroundWater, v. 26, no. 3, p. 333–338. Wells, M. B., 1980, Reflections on the evolution of algorithmic language, in Metropolis, N., Howlett, J., and Rota, G.-C., eds., A history of Computing in the Twentieth Century: Academic Press, New York, p. 275–287. Whiston, W., 1719, The longitude and latitude found by the inclinatory or dipping needle: Printed for author, London, 32 p. Whiston, W., 1721, The longitude and latitude found by the inclinatory or dipping needle; Wherein the laws of magnetism are also discover'd. To which is prefix'd, an historical preface; and to which is subjoin'd, Mr. Robert Norman's new attractive, or account of the first invention of the dipping needle: Senex and Taylor, London, xviii p. C115 p. Whitten, E. H. T., 1959, Composition trends in a granite: modal variation and ghost stratigraphy in part of the Donegal granite, Eire: Jour. Geophys. Research, v. 64, no. 7, p. 835–848. Whitten, E. H. T., 1961, Quantitative areal modal analysis of granitic complexes: Geol. Soc. America Bull., v. 72, no. 9, p. 1331–1360. Whitten, E. H. T., 1963, A surface-fitting program suitable for testing geological models which involve areally-distributed data: Tech. Rept. no. 2 of ONR Task no. 389-135, Geography Branch, Office of Naval Research. Northwestern Univ., Evanston, Illinois, 56 p. Whitten, E. H. T., 1964, Best mathematical model for mapped variables: Gold variability at Virginia Mine, South Africa (abst.), in Program with Abstracts, Miami Meeting, Geol. Soc. America, p. 223. Whitten, E. H. T., 1966, The general linear equation in prediction of gold content in Witwatersrand rock, South Africa: South African Inst. Mining and Metallurgy, v. 66, no. 3, p. 124–165. Whitten, E. H. T., 1974, Orthogonal-polynomial contoured trendsurface maps for irregularly-spaced data: Tech. Rept. 2. U.S. Army Research office-Durham grant DA-ARO-D-31-124-72-G54. Northwestern Univ., Evanston, Illinois, p. 1–24. Whitten, E. H. T., and Boyer, R. E., 1964, Process-response models based on heavy-mineral content of the San Isabel Granite, Colorado: Geolo. Soc. America Bull., v. 75, no. 9, p. 841–862. Will, T. M., and Powell, R., 1991, A robust approach to the calculation of paleostress fields from fault plane data: Jour. Structural Geology, v. 13, no. 7, p. 813–821. Williamson, J. H., 1968, Least-squares fitting of a straight line: Can. Jour. Physics, v. 44, p. 1097–1086. Winchell, H., and Tilling, R., 1960, Regressions of physical properties on the composition of clinopyroxenes. I. Lattice constants: Am. Jour. Science, v. 258, no. 8, p. 529–547. Wold, H. O. A., 1938, A study in the analysis of stationary time series: Alnquist and Winsell, Uppsala, 214 p. Wold, H.O. A., 1949,Alarge sample test of moving averages: Jour. Royal Stat. Society, v. B11, p. 297–305. Wold, S., 1974, Spline functions in data analysis: Technometrics, v. 16, no. 1, p. 1–11. Woolf, B., 1951, Computation and interpetation of multiple regressions: Jour. Royal Stat. Society, v. B13, no. 1, p. 100–119. York, D., 1966, Least-squares fitting of a straight line: Can. Jour. Physics, v. 44, p. 1079–1086. York, D., 1967, The best isochron: Earth and Planetary Science Letters, v. 2, no. 5, p. 479–482. York, D., 1969, Least-squares fitting of a straight line with correlated errors: Earth and Planetrary Science Letters, v. 5, no. 5, p. 320–324. Yule, G. U., 1897, On the theory of correlation: Jour. Royal Stat. Society, v. 60, p. 812–854. Zhou, D., 1987, Robust statistics and geochemical data analysis: Math. Geology, v. 19, no. 3, p. 207–218.

bne: Кстати, работа Krumbein-Monk по предсказанию проницаемости по гранулометрии у него выпала

bne: Получил автореферат на отзыв БЕЛЯНУШКИНА МАРИЯ СЕРГЕЕВНА ПРОГРАММНО-АЛГОРИТМИЧЕСКОЕ ОБЕСПЕЧЕНИЕ КЛАСТЕРИЗАЦИИ И ВИЗУАЛИЗАЦИИ МНОГОМЕРНЫХ ДАННЫХ СЕЙСМОРАЗВЕДКИ И ГИС Специальность 25.00.10 – геофизика, геофизические методы поисков полезных ископаемых АВТОРЕФЕРАТ диссертации на соискание ученой степени кандидата технических наук МОСКВА 2009 =========================== Есть за что и поругать чуток и похвалить Шапочно не увидел ссылок на Бонгарда и Губермана Аналогично - ссылок на работы по смесям (а штука это полезная и я это в книжке с Элланским писал и показывал давным-давно) и работы по сетям Кохонена С остальным можно разбираться

bne: В связи с авторефератом интересная мне ссылка (там есть и хитрости про двумерную медиану и кластеры и прочие игры) http://valis.cs.uiuc.edu/~sariel/papers/ А работа интересная, жаль автор немного и шире и дальше в прошлое методов не пытается посмотреть Применения также интересны (и программные и прикладные) И еще http://users.cis.fiu.edu/~giri/publications.html И кроме того X-means: Extending K-means with Efficient Estimation of the Number of Clusters (2000) Dan Pelleg, Andrew Moore Abstract A K-means tutorial. Despite its popularity for general clustering, k-means suffers three major shortcomings; it scales poorly computationally, the number of clusters K has to be supplied by the user, and the search is prone to local minima. We propose solutions for the first two problems, and a partial remedy for the third. Building on prior work for algorithmic acceleration that is not based on approximation, we introduce a new algorithm that efficiently, searches the space of cluster locations and number of clusters to optimize the Bayesian Information Criterion (BIC) or the Akaike Information Criterion (AIC) measure. The innovations include two new ways of exploiting cached sufficient statistics and a new very efficient test that in one k-means sweep selects the most promising subset of classes for refinement. This gives rise to a fast, statistically founded algorithm that outputs both the number of classes and their parameters. Experiments show this technique reveals the true number of classes in the underlying distribution, and that it is much faster than repeatedly using accelerated k-means for different values of K.

bne: Отзыв на автореферат диссертации Белянушкиной Марии Сергеевны ПРОГРАММНО-АЛГОРИТМИЧЕСКОЕ ОБЕСПЕЧЕНИЕ КЛАСТЕРИЗАЦИИ И ВИЗУАЛИЗАЦИИ МНОГОМЕРНЫХ ДАННЫХ СЕЙСМОРАЗВЕДКИ И ГИС представленной на соискание ученой степени кандидата технических наук по специальности 25.00.10 – геофизика, геофизические методы поисков полезных ископаемых Тема диссертации М.С.Белянушкиной относится к одному из актуальнейших и интенсивно развиваемых направлений современной мировой науки – визуальным методам Data Mining – и представляет собой интересную попытку авторской адаптации развиваемых в этой области методов применительно к задачам сейсморазведки и ГИС. Судя по автореферату, автор всерьез подошел к решению этой задачи и провел полезную и интересную работу. Невзирая на сотни публикаций в области построения связей между различными сейсмическими атрибутами и геологическими параметрами, предлагаемые решения представляют значимый интерес для специалистов в этой сфере. Первая глава содержит обзор работ, посвященных теме диссертации. Приводимый обзор интересен, но, судя по автореферату, автор исходит, скорее, из изучения достижений математиков (в видении, близком к излагаемому в книгах В.Н. Вапника и ряде современных публикаций), чем из анализа исследований по тематике в области геологии и геофизики. Поскольку поднятая тема в настоящее время в связи возрастанием количества исследований на глазах становится все менее обозримой, полный обзор работ представлял бы собой отдельную и серьезную задачу. Тем не менее, досадно, когда из перечисления авторитетов выпадают такие пионеры в этой области как наши соотечественники М.М.Бонгард и Ш.А.Губерман, которые, по сути, открывали эту проблематику для задач геофизики в 60-е годы прошлого века. Забывая приоритет предшественников, мы невольно занижаем остатки авторитета наших современников. В этой же главе дано также краткое описание ряда пакетов, с применением которых возможна реализация визуализации в различных приложениях. Во второй главе дано описание разработанных автором алгоритмов кластеризации, позволяющих решить несколько важных прикладных задач. Приведено описание созданной автором оригинальной системы тестов для проверки и оценки результатов работы алгоритмов кластеризации. Предложенная система тестов позволяет сделать выводы о вероятности правильного разбиения ЧЕГО НА ЧТО??? , что является, несомненно, новым элементом , а также о зависимости корректной работы алгоритмов от соотношения плотности частиц в кластерах и плотности частиц в исследуемом поле. Приведены краткие сведения о библиотеке OpenGL, которая недостаточно известна геофизикам и весьма полезна при развитии систем графической визуализации. В четвертой главе описаны практические применения разработанных алгоритмов и программ. В этой главе приводится описание результатов кластеризации сейсмических данных: карт сейсмических фаций, кубов кластеризации, разработана методика расчета куба кластеризации. Приведены также примеры визуализации интерпретации различных ГИС: выделения аномальных интервалов по комплексу каротажа на месторождении с нетрадиционным коллектором, интерпретации каротажа горизонтальных скважин, интерпретации СО-каротажа. В целом по работе можно высказать одно пожелание и, пожалуй, не к диссертанту. Судя по названию диссертации и тексту автореферата, М.С.Белянушкина берется за решение практически глобальных и неподъемных даже в докторской диссертации задач и потому временами создается впечатление неполноты раскрытия чересчур претенциозно сформулированных тем. Так, по сути, в диссертации, конечно, разрабатывается лишь один из классов алгоритмов кластеризации, оптимальный для специальных классов задач. При рассмотрении в этом же ключе становится понятно, почему и набор тестов, сделанный профессионально и интересно, все же специализирован. При таком замахе становится непонятно, почему автор не анализирует результаты по другим алгоритмам, большой набор которых имеется, в частности, в цитируемой в диссертации книге Hartigan’а. В области кластеризации есть свои школы и вкусы. Например, автор отзыва давно и не без успеха использует для решения сходных задач методы разделения смесей распределений и методы размытых средних. И, конечно, когда подобные разработки не проанализированы и даже не упомянуты при столь всеобъемлющем названии, это вызывает досаду. Cтоль же необъятно широко сформулирован еще более широкий ракурс проблематики – визуализация. Однако было бы неверно отнести пожелание привести в соответствие текст работы с ее название и названием ее разделов к диссертанту лично. Если отвлечься от подобных недоразумений, работа производит впечатление интересной и добротно выполненной во всех ее пунктах, которые носят конкретный и содержательный характер (от конкретных алгоритмов и программ до конкретных приложений). В этой связи хочется поздравить и автора, и научного руководителя с интересным исследованием. В целом работа, если судить по автореферату, соответствует требованиям ВАК по указанной специальности. На основании изложенного считаю, что Мария Сергеевна Белянушкина заслуживает присуждения ей искомой ученой степени кандидата технических наук по специальности 25.00.10 — «Геофизика, геофизические методы поисков полезных ископаемых». Главный петрофизик ЗАО «Пангея» к.т.н. Б.Н. Еникеев Подпись Б.Н. Еникеева заверяю Генеральный директор ЗАО «Пангея» В.В.Колесов

Andrew: Борис Николаевич, может у вас есть емейл этой самой Белянушкиной Марии Сергеевны? Не поделитесь? Интересно бы спросить у неё, может даст почитать.

bne: Я получал реферат письмом от совета и отправлял скан текстом Лухминского boris.lukhminsky "ЭТО_СОБАКА_ЗНАЧИТ" gmail.com А по жизни вроде ничего экстремально интересного, но работа, а не халтура (как это часто просматривается) По большому счету она IMHO не самый удачный метод для своих задач выбрала (по рисунку у нее эллипсоиды) Да и могла бы на пристойных привычных тестах (начиная от ИРИСОВ ФИШЕРА) провериться

bne: ISSN 2070-5379 Neftegasovaa geologia. Teoria i practika (RUS) URL: http://www.ngtp.ru 1 Лабковский В.В. Институт нефтегазовой геологии и геофизики им. А.А. Трофимука СО РАН (ИНГГ СО РАН), Новосибирск, Россия, lapkovskiivv@ipgg.sbras.ru Шмелев Н.Е. Новосибирский государственный университет (НГУ), Новосибирск, Россия, shmelevne@ipgg.sbras.ru Лунёв Б.В. Институт нефтегазовой геологии и геофизики им. А.А. Трофимука СО РАН (ИНГГ СО РАН), Новосибирск, Россия, lunevbv@ipgg.sbras.ru ОБОБЩЁННЫЙ ОБРАЗ СКВАЖИНЫ ПО КАРОТАЖНЫМ ДАННЫМ: ТЕХНОЛОГИЯ СОЗДАНИЯ И ПРИМЕНЕНИЯ По аналогии с методом обобщенного фотопортрета предлагается создание обобщенного обрат определенного множества скважин по их каротажным данным. Совмещение отдельных скважин выполняется на основе корреляционной модели, которая позволяет установить в них наборы стратиграфически эквивалентных точек. Построение стратиграфической модели рассматривается как решение многомерной оптимизационной задачи. Обобщенный образ скважины позволяет количественно оценивать типичные характеристики свит и пластов, анализировать индивидуальные отклонения, свойственные отдельным районам или скважинам. На основании этого подхода возможно прогнозирование распределения свойств для проектируемых скважин и создание трехмерных моделей (кубов) свойств изучаемых объектов. Ключевые слова: автоматическая корреляция скважин, каротажные данные, метод обобщенного портрета, обобщённый образ скважины, прогнозирование разреза, кубы свойств.

bne: И вроде как и интересно, только странно даже по ссылкам Уж упоминая Жековского нужно было бы и Губермана и даже Гутмана упомянуть Не понимаю такого отношения Тем более, что приём с центрированными средними применялся думаю не десятки, а сотни раз Да и сама по себе идея сравнения в исходном пространстве требует в частности предварительной нормировки И с предлагаемым функционалом трудно не спорить

bne: Data Mining - Artificial neural networks; Bayesian networks; Heuristic methods; Case-based reasoning; Computational models of human learning (Cooperative learning, Unsupervised learning); Optimization methods; Decision and induction methods; Evolutionary computation; Grammatical inference; Incremental and on-line learning; Information retrieval and learning; Knowledge acquisition and learning; Data pre - and post-processing; Data visualization; Statistical pattern recognition and analysis; Bootstrap and randomization; Causal modeling; Decision analysis; Exploratory data analysis; Knowledge-based analysis; Classification, projection, regression, optimization clustering; Data cleaning; Model specification, selection, estimation; Reasoning under uncertainty; Uncertainty and noise in data; Spatial Analysis; Space-Time Analysis; Discrete Events Analysis, Network Analysis; Time series forecasting и т.д. Text Mining - Bayesian Models, Concept Decomposition, Orthogonal Decomposition, Probabilistic Models, Vector Space Models, Latent Semantic Indexing, Graph-based Models, Text Streaming Models, Clustering, Factor Analysis, Visualization Techniques, Metadata Generation, Information Extraction, Text Classification, Text Purification, Text Segmentation, Text Summarization, Query Structures, Trend Detection, Distributed Storage and Retrieval. Взято по ссылке: http://www.isras.ru/index.php?page_id=988 Там вообще много интересного широкими мазками прорисовано По хорошему грех каротажу и петрофизике (начинавших игры с распознованием) отставать от вполне гуманитарных дисциплин И не только по анализу данных, но и по анализу текстов

bne: Оригинал статьи Рональда Фишера (в которой заложены основы кластерного анализа) 1936 год http://digital.library.adelaide.edu.au/coll/special//fisher/138.pdf Фото Рональда Фишера А вот про дендрограмму я пока не нашел

Василий: Эх жаль что на английском, долго разбираться.(

bne: Про статью не часто вспоминают, но "Ирисы Фишера" из этой статьи - самый распространенный тест для кластерного анализа

bne: Интересно видит кое-кто историю распознавания образов In 1936, R. A. Fisher suggested the first algorithm for pattern recognition (Fisher 1936). Aronszajn (1950) introduced the ‘Theory of Reproducing Kernels’. In 1957 Frank Rosenblatt invented a linear classifier called the perceptron (the simplest kind of feedforward neural network), see Rosenblatt (1962). Vapnik and Lerner (1963) introduce the Generalized Portrait algorithm (the algorithm implemented by support vector machines is a nonlinear generalization of the Generalized Portrait algorithm). Aizerman, Braverman and Rozonoer (1964) introduced the geometrical interpretation of the kernels as inner products in a feature space. Vapnik and Chervonenkis (1964) further develop the Generalized Portrait algorithm. Cover (1965) discussed large margin hyperplanes in the input space and also sparseness. Similar optimisation techniques were used in pattern recognition by Mangasarian (1965). The use of slack variables to overcome the problem of noise and nonseparability was introduced by Smith (1968). Duda and Hart (1973) discuss large margin hyperplanes in the input space. The field of ‘statistical learning theory’ began with Vapnik and Chervonenkis (1974) (in Russian). SVMs can be said to have started when statistical learning theory was developed further with Vapnik (1979) (in Russian). Wapnik and Tscherwonenkis (1979) wrote a German translation of Vapnik and Chervonenkis?s 1974 book. Vapnik (1982) wrote an English translation of his 1979 book. See also the PhD thesis by Hassoun (1986) for related early work. Several statistical mechanics papers (for example Anlauf and Biehl (1989)) suggested using large margin hyperplanes in the input space. Poggio and Girosi (1990) and Wahba (1990) discuss the use of kernels. Bennett and Mangasarian (1992) improved upon Smith’s 1968 work on slack variables. SVMs close to their current form were first introduced with a paper at the COLT 1992 conference (Boser, Guyon and Vapnik 1992). In 1995 the soft margin classifier was introduced by Cortes and Vapnik (1995); in the same year the algorithm was extended to the case of regression by Vapnik (1995) in The Nature of Statistical Learning Theory. The papers by Bartlett (1998) and Shawe-Taylor, et al. (1998) gave the first rigorous statistical bound on the generalisation of hard margin SVMs. Shawe-Taylor and Cristianini (2000) gave statistical bounds on the generalisation of soft margin algorithms and for the regression case. Links Support Vector Learning: Bernhard Sch?pf Bibliography AIZERMAN, M. A., E. M. BRAVERMAN, and L. I. ROZONOER, 1964. Theoretical foundations of the potential function method in pattern recognition learning. Automation and Remote Control, 25, 821–837. ANLAUF, J. K., and M. BIEHL, 1989. The adatron: An adaptive perceptron algorithm. Europhysics Letters, 10(7), 687–692. ARONSZAJN, N., 1950. Theory of reproducing kernels. Transactions of the American Mathematical Society, 68(3), 337–404. BARTLETT, Peter L., 1998. The sample complexity of pattern classification with neural networks: The size of the weights is more important than the size of the network. IEEE Transactions on Information Theory, 44(2), 525–536. BENNETT, Kristin P., and O. L. MANGASARIAN, 1992. Robust linear programming discrimination of two linearly inseparable sets. Optimization Methods and Software, 1, 23–34. BOSER, Bernhard E., Isabelle M. GUYON, and Vladimir N. VAPNIK, 1992. A training algorithm for optimal margin classifiers. In: COLT ’92: Proceedings of the Fifth Annual Workshop on Computational Learning Theory. New York, NY, USA: ACM Press, pp. 144–152. CORTES, Corinna, and Vladimir VAPNIK, 1995. Support-vector networks. Machine Learning, 20(3), 273–297. COVER, Thomas M., 1965. Geometrical and statistical properties of systems of linear inequalities with applications in pattern recognition. IEEE Transactions on Electronic Computers, 14(3), 326–334. DUDA, Richard O., and Peter E. HART, 1973. Pattern Classification and Scene Analysis. New York: John Wiley & Sons Inc. FISHER, R. A., 1936. The use of multiple measurements in taxonomic problems. Annals of Eugenics, 7, 111–132. HASSOUN, M. H., 1986. Optical Threshold Gates and Logical Signal Processing. Ph. D. thesis, Wayne State University, Detroit, USA. MANGASARIAN, O. L., 1965. Linear and nonlinear separation of patterns by linear programming. Operations Research, 13(3), 444–452. POGGIO, Tomaso, and Federico GIROSI, 1990. Networks for approximation and learning. Proceedings of the IEEE, 78(9), 1481–1497. ROSENBLATT, Frank, 1962. Principles of Neurodynamics: Perceptrons and the Theory of Brain Mechanisms. Washington DC: Spartan Books. SHAWE-TAYLOR, John, et al., 1998. Structural risk minimization over datadependent hierarchies. IEEE Transactions on Information Theory, 44(5), 1926–1940. SHAWE-TAYLOR, John, and Nello CRISTIANINI, 2000. Margin distribution and soft margin. In: Alexander J. SMOLA, et al., eds. Advances in Large Margin Classifiers. The MIT Press, pp. 349–358. SMITH, F. W., 1968. Pattern classifier design by linear programming. IEEE Transactions on Computers, C-17(4), 367–372. VAPNIK, V., 1979. Estimation of Dependences Based on Empirical Data [in Russian]. Moscow: Nauka. VAPNIK, Vladimir, 1982. Estimation of Dependences Based on Empirical Data. Springer Verlag. VAPNIK, V., and A. CHERVONENKIS, 1964. A note on one class of perceptrons. Automation and Remote Control, 25. VAPNIK, V., and A. LERNER, 1963. Pattern recognition using generalized portrait method. Automation and Remote Control, 24, 774–780. VAPNIK, Vladimir N., 1995. The Nature of Statistical Learning Theory. Springer-Verlag New York, Inc. VAPNIK, V. N., and A. Ya. CHERVONENKIS, 1974. Teoriya raspoznavaniya obrazov: Statisticheskie problemy obucheniya. (Russian) [Theory of pattern recognition: Statistical problems of learning]. Moscow: Nauka. WAHBA, Grace, 1990. Spline Models for Observational Data. Volume 59 of CBMS-NSF Regional Conference Series in Applied Mathematics. Philadelphia, PA, USA: SIAM: Society for Industrial and Applied Mathematics. WAPNIK, W. N., and A. J. TSCHERWONENKIS, 1979. Theorie der Zeichenerkennung. (German) [Theory of pattern recognition]. Berlin: Akademie-Verlag. http://www.svms.org/history.html

bne: Giulio D. Agostini, “Bayesian Reasoning in Data Analysis: A Critical Introduction” World Scientific Publishing Company | 2003 | ISBN: 9812383565 | 329 pages | This book provides a multi-level introduction to Bayesian reasoning (as opposed to "conventional statistics") and its applications to data analysis. The basic ideas of this "new" approach to the quantification of uncertainty are presented using examples from research and everyday life. Applications covered include: parametric inference; combination of results; treatment of uncertainty due to systematic errors and background; comparison of hypotheses; unfolding of experimental distributions; upper/lower bounds in frontier-type measurements. Approximate methods for routine use are derived and are shown often to coincide — under well-defined assumptions! — with "standard" methods, which can therefore be seen as special cases of the more general Bayesian methods. In dealing with uncertainty in measurements, modern metrological ideas are utilized, including the ISO classification of uncertainty into type A and type B. These are shown to fit well into the Bayesian framework.

bne: Statistical Pattern Recognition, 2nd Edition Publisher: Wiley | ISBN: 0470845139 | edition 2002 | 534 pages Statistical pattern recognition is a very active area of study and research, which has seen many advances in recent years. New and emerging applications - such as data mining , web searching, multimedia data retrieval, face recognition, and cursive handwriting recognition - require robust and efficient pattern recognition techniques. Statistical decision making and estimation are regarded as fundamental to the study of pattern recognition. Statistical Pattern Recognition, Second Edition has been fully updated with new methods, applications and references. It provides a comprehensive introduction to this vibrant area - with material drawn from engineering, statistics, computer science and the social sciences - and covers many application areas, such as database design, artificial neural networks, and decision support systems. * Provides a self-contained introduction to statistical pattern recognition. * Each technique described is illustrated by real examples. * Covers Bayesian methods, neural networks, support vector machines, and unsupervised classification. * Each section concludes with a description of the applications that have been addressed and with further developments of the theory. * Includes background material on dissimilarity, parameter estimation, data, linear algebra and probability. * Features a variety of exercises, from 'open-book' questions to more lengthy projects. The book is aimed primarily at senior undergraduate and graduate students studying statistical pattern recognition, pattern processing, neural networks, and data mining, in both statistics and engineering departments. It is also an excellent source of reference for technical professionals working in advanced information development environments. http://www.petrophysics.borda.ru/?1-5-0-00000107-011

bne: Aspects of Multivariate Statistical Analysis in Geology by E. Savazzi and R.A. Reyment Publisher: Elsevier Science | November 1, 1999 | ISBN: 0444504125 | Pages: 250 The book presents multivariate statistical methods useful in geological analysis. The essential distinction between multivariate analysis as applied to full-space data (measurements on lengths, heights, breadths etc.) and compositional data is emphasized with particular reference to geochemical data. Each of the methods is accompanied by a practically oriented computer program and backed up by appropriate examples. The computer programs are provided on a compact disk together with trial data-sets and examples of the output. An important feature of this book is the graphical system developed by Dr. Savazzi which is entitled Graph Server. Geological data often deviate from ideal statistical requirements. For this reason, close attention has been paid to the analysis of data that contain atypical observations.

bne: Смотрел отчет по Ямбургу Ребята там применили кластерный анализ из САПФИРА Коллега Ермаков как всегда проявляет себя самородком Применяется потенциальная функция вводимая как обратная степенная и уровень отсечки Естественно что это приводит к результатам близким к к-средним

bne: Kernels For Structured Data (Series in Machine Perception & Art Intelligence) This book provides a unique treatment of an important area of machine learning and answers the question of how kernel methods can be applied to structured data. Kernel methods are a class of state-of-the-art learning algorithms that exhibit excellent learning results in several application domains. Originally, kernel methods were developed with data in mind that can easily be embedded in a Euclidean vector space. Much real-world data does not have this property but is inherently structured. An example of such data, often consulted in the book, is the (2D) graph structure of molecules formed by their atoms and bonds. The book guides the reader from the basics of kernel methods to advanced algorithms and kernel design for structured data. It is thus useful for readers who seek an entry point into the field as well as experienced researchers. Contents: Why Kernels for Structured Data?; Kernel Methods in a Nutshell; Kernell Design; Basic Term Kernels; Graph Kernels.

bne: EAGE 2009 Amsterdam Q046 Evaluating Different Approaches of Permeability Modeling in Heterogeneous Carbonate Reservoirs F. Khoshbakht* (RIPI), M. Mohammadnia (RIPI), A.M. Bagheri (RIPI), A.A. RahimiBahar (RIPI) & Y. Beiraghdar (RIPI) SUMMARY Permeability, the ability of rocks to flow hydrocarbons is directly determined in the laboratory on cores taken from the reservoir. Due to high cost associated with coring, many empirical models, statistical methods and intelligence techniques were suggested to establish robust relationships between permeability and various easy to obtain and frequent data such as wireline logs. This study launched to put different approaches of permeability modeling into practice to predict permeability in a heterogeneous carbonate reservoir (Fahliyan formation in SW Iran) and compare results in order to determine the optimal approach for utilizing in this formation. Considered methods divide into four groups; a) empirical models (Timur and dual water), b) regression analysis (simple and multiple), c) clustering methods (MRGC, SOM, DC & AHC) and d) artificial intelligence techniques (ANN, fuzzy logic and neuro-fuzzy). This study shows that clustering techniques predict permeability in a heterogeneous carbonate better than other examined approaches. Among four assessed clustering methods, SOM performs best and could handle dimensionality and complexity of input data sets. Artificial intelligence techniques are average in modeling permeability, in addition empirical equations and regression techniques are not capable for predicting permeability in studied heterogeneous carbonate reservoir.

B_N_E_1: Fred Ramsey, Daniel Schafer, "The Statistical Sleuth: A Course in Methods of Data Analysis" Duxbury Press | 2001-10-05 | ISBN: 0534386709 | 768 pages | STATISTICAL SLEUTH is an innovative treatment of general statistical methods, taking full advantage of the computer, both as a computational and an analytical tool. The material is independent of any specific software package. In "The American Statistician" (February 2000, Vol. 54, No. 1), George Cobb commented, "What is new and different about Ramsey and Schafer's book, what makes it a 'larger contribution,' is that it gives much more prominence to modeling and interpretation of the sort that goes beyond the routine patterns." His students did "substantially better" on term papers based on the analysis of data. In the book, the focus is on a serious analysis of real case studies; on strategies and tools of modern statistical data analysis; on the interplay of statistics and scientific learning; and on the communication of results. With interesting examples, real data, and a variety of exercise types (conceptual, computational, and data problems), the authors get students excited about statistics.

bne: Ronald P. Cody, "Cody's Data Cleaning Techniques Using SAS Software" SAS Publishing | 1999-11-29 | ISBN: 1580256007 | 226 pages | The key to ensuring accurate data is having clean data. This book develops and describes data cleaning programs and macros. You can use many of the programs and macros that are provided, as is, or you can modify them for your own special data cleaning tasks. Ron has carefully explained and documented each of the programs and macros, thus providing you with SAS programming instruction on an intermediate-to-advanced level. Topics presented include validation checks on character data, numeric data, missing values, and date values; searching for duplicate records; working with multiple files; double entry and verification using the COMPARE procedure; and SQL solutions and using validation data sets. Written in Ron's signature informal, tutorial style, this book gives anyone who manages data thoroughly documented, step-by-step instructions for the development of data cleaning programs and macros. Supports releases 6.12 and higher of SAS software. Тема очистки данных от загрязнений - одна из важнейших (сам этим занят регулярно) Кое-что для этого есть и в ModERn, но в основном по разделению техст-числа и интерактивному поиску грубых ошибок Хорошо бы изучить рекомендации автора - может что полезное найдется ;-)

bne: D. M. Titterington, "Statistical Analysis of Finite Mixture Distributions" John Wiley & Sons | 1986 | ISBN: 0471907634 | 254 pages In this book, the authors give a complete account of the applications, mathematical structure and statistical analysis of finite mixture distributions. This area of statistics is important to a range of disciplines, and it's methodology is attracting interest from researchers in the fields in which it can be applied. Applications include sedimentology, fisheries research, medicine, remote sensing and economics. More indirect applications include outlier models, density estimation, Bayesian and empirical Bayes analysis and robustness studies. All these areas of application are discussed in detail, before the related mathematical concepts are introduced. Concepts discussed include identifiability and information, and there is a substantial account of the inferential problems associated with using data from a mixture. A particularly awkward problem tackled here is that of deciding how many components the underlying mixture contains. Because of estimation difficulties, and to cope with applications in engineering, the authors have developed approximate sequential methods, these are described in the final chapter.

viking23: Часто расматриваю иногранные книги, очень красиво выглядят. Так и хочется купить и отнести в свою коллекцию. Вопрос вам, Есть ли список литературы, экземпляры которой имеют фундаментальное значение(петрофизика, ГИС методы)? Хочу по возможности создать свою бумажную библиотеку, зарубужного происхождения. Но нет возможности смотреть всё подряд. Ещё интересно про оцифровку книг, Гугл. Там хоть кусочками можно посмотреть кое какие- издания. Но насколько я понял, гуглу не дадут отсканировать всё ценные книги:)Правда ли?

bne: 1) Я часто публикую аннотации книг доступных в электронном виде в интернете Ссылки теперь не даю (даже в закрытом разделе), поскольку Yandex из-за этого исключает из результатов поиска 2) Есть средства считывания из Google Books Все неисключенные страницы можно считать и несколькими способами (в *.PDF или *.PNG) Но за это GoogleBooks перекрывает доступ 3) Есть еще Springer (все более открывающийся) и Elsiever (в свое время у РосНефти был туда доступ в Scorpions - так запомнилось) Думаю что все ограничения на книги - вопрос времени А вот труды конференцией (типа SPE и SPWLA) имеют тенденцию к большему закрытию Тем не менее представляется, что сети файлообмена (в первую очередь приверженцы ислама) станут противодействовать оследней тенденции В этой связи на первое место могут начать выходить авторизованные обзоры (к чему у меня лично есть склонность) Знание нужно более адекватное и переваренное Но с его публикацией также проблемы с авторским правом (правда в обратную сторону) - труд это затратный и легко переприсваемый (что ослабляет мотивацию если не параноик) Как быстро это пойдет - неясно

viking23: у нас в институте появился доступ к SPE, это радует

bne: Увкерен, что мне нашего босса не раскрутить У меня есть тлько до 1997 года и с сотню других отдельных выпусков

viking23: Странно, но сейчас зашел и оказывается скачать статьи от туда можно только за деньги, при этом с ресурса one-petro они всё скачиваются бесплатно(с статьи из последних номеров),

bne: Люди делают бизнес на ровном месте Продают чужой труд и покупатели платят за чисто информационно-организационные услуги И все по причине того, что нет энтузиастов готовых депонировать публикации Скажем у социологов есть огромныне хранилища (на лесятки тысяч статей) и все на халяву

viking232: Получается нефтяная отрасль в этом схожа на автомобильную тему. Всё что с ней связанно. Либо за огромные деньги реализуется, либо вообще не реализуется. Я сейчас собираюсь создавать базу данных публицаций, но пока не нашел подходящей программы. Единственное более менее подходящие http://cookrecepts.narod.ru/ она бесплатная и можно писать в одной графе сколько угодно строк. Нет ли каких-то более подходящих программ для создания базы публикаций? //viking23

bne: Я не в теме Я таких вещей побаиваюсь и предпочитаю работать с файлами Если содержимое публикаций не растры, то использую Copernic Вообще же генерация создания HTML файла из названий (если речь не идет о десятках тысяч и более) на современных персоналках проблемы не составляет Я ориентирован на подобные простые решения Если речь идет о внешних хранилищах. то заботу могут взять на себя готовые движки Я побаиваюсь программ с отсутствующими исходниками и сомнительными обязательствами

viking23: нет тут вопрос в нахождении нужной публикации как бы она не называлась. Например можно привязать к какому-то файлу поля, имя файла, название, ключевые слова. и создать из названий фалов и полей -базу данных. Так наверно реализован движок onepetro или seg сборник статьей по результатом технической научной на диске написанный вроде на Flash. сейчас для меня актуальна задача классификации статей). статей, (pdf. word, djvu) а не файлов. Менеджеры файлов немного не то.

bne: Надо и времянки создавать и разгребать и имитировать персональные описания и оценки Вопрос в % сил который на это естественно тратить

viking23: сколько профита(пользы) от этого дела. мне эстетически доставляет удовольствие от созерцания структурированной базы данных. Одна проблема, что с каждым годом расширяется горизонт познаний и появляются новые отрасли или категории по которым надо разбивать новый, и возможно старые). общем если интересно, есть программка отечественная бесплатная(работает под {H под W&x64 падает): http://catbook.narod.ru/ она позволяет систематизировать книги (электронные версии). с записью информации, представления в древовидном виде, но автоматически сканировать каталог, она не может. Я думаю сейчас в этом направлении)

БНЕ_Home: Включу в план работ Сейчас из Уфы приваез кучу дополнительных задач

viking23: то есть направление должно быть такое, чтобы можно было и посмотреть книги в каталоге, и про сканировать дирректорию или несколько вложенных. Я уже переименовал 300 своих статей. теперь осталось дело за аппаратной частью. Суть в том что иногда одного автора в заголовке не достаточно. хотя может это идеально, И второе что некоторые ФИО странно пишутся. macDonald Al-Faraby, T-f Wond.как их сокращать mDonald MDonald? aFaraby WongTf в общем пока такие забавности.

bne: В принципе есть свои достоинства в Copernic Но уж больно серьезно он подходит к процессу Долго индексный файл составляет и много мест он забирает

bne: я не могу найти ссылку на обсуждаемую тему про организацию научных статей на компьюторе. Добавьте пожалуйста туда. Я сделал программу которой можно немного это дело систематизировать, она позволяет делать разбивку по 5 авторам, году, 5 ключевым словам, журналу, и уникальному имени(у меня оригинальное название файла, чтобы сравнивать со свеже скаченным)-это категории. Они отражаются в названии файлов через разделитель_так. Внутри катогорий разделитель запятая. Также нашел смежную тему под данному направлению http://www.rsdn.ru/forum/life/3388625.hot.aspx Может быть интересно =============== PS Автор просил в личке положить этот текст в ранее начатую тему Таки удалось найти ;-)

bne: Hybrid Methods in Pattern Recognition Publisher: World Scientific Publishing Company | ISBN: 9810248326 | edition 2002 | 338 pages Collection of articles describing recent progress in this emerging field. Covers topics such as the combination of neural nets with fuzzy systems or hidden Markov models, neural networks for the processing of symbolic data structures, hybrid methods in data mining, and others.

bne: "Robust Regression and Outlier Detection (Wiley Series in Probability and Statistics)" Wiley (October 1987) | English | 0471852333 | 352 pages Provides an applications-oriented introduction to robust regression and outlier detection, emphasising "high-breakdown" methods which can cope with a sizeable fraction of contamination. Its self-contained treatment allows readers to skip the mathematical material which is concentrated in a few sections. Exposition focuses on the least median of squares technique, which is intuitive and easy to use, and many real-data examples are given. Chapter coverage includes robust multiple regression, the special case of one-dimensional location, algorithms, outlier diagnostics, and robustness in related fields, such as the estimation of multivariate location and covariance matrices, and time series analysis.

bne: Herbert K. H. Lee «Baysian Nonparametrics via Neural Networks (ASA-SIAM Series on Statistics and Applied Probability)» Publisher: Society for Industrial and Applied Mathematics | Edition: 2004 | ISBN: 0898715636 | PDF | 104 pages Bayesian Nonparametrics via Neural Networks is the first book to focus on neural networks in the context of nonparametric regression and classification, working within the Bayesian paradigm. It discusses neural networks in a statistical context, an approach in contrast to existing books, which tend to treat neural networks as a machine learning algorithm instead of a statistical model. Once this underlying statistical model is recognized, other standard statistical techniques can be applied to improve the model. The Bayesian approach allows better accounting for uncertainty. This book covers uncertainty in model choice and ways to deal with this issue, exploring ideas from statistics and machine learning. An analysis on the choice of prior and new noninformative priors is included, along with a substantial literature review. Written for statisticians using statistical terminology, this book will lead statisticians to an increased understanding of the neural network model and its applicability to real-world problems.

bne: Давно я не смотрел литературы на сей счет (история моего вхождения в тему по ссылке http://petrophysics.borda.ru/?1-7-0-00000013-006-0-0 ), но похоже, что уже перезрел Суть идеи в оптимальной комбинации решений полученных альтернативными способами (причем чем более разнообразныит, но желательно равноточными - тем лучше)

bne_mumbai2: Пришлось работать с весьма сложным объектом Поэтому разделил презентацию на две В первой описание подхода и петрофизики, а во второй - проблемы При этом проблемыв я структурировал в духе ЛОГИКИ Милля (по смягчению или отказу от допущений) Получился неплохой фрейм-канва (намного более широкий чем вопросы и замечания) Проблемы удались много лучше (увы так чаще всего и бывает) В итоге проблемы работы утонули в широте и частностях проблем объекта Думаю это правильно не только тут, но и вообще Чем лучше понимание сложностей тем больше оснований для полноценного диалога с заказчиком и уважение ксамому себе

bne: I.А. Козачок Моделi петрофiзичних зв’язкiв у обернених задачах двозондового нейтрон-нейтронного каротажу (Представлено академiком НАН України В. I. Старостенком) Запропоновано й дослiджено моделi багатовимiрних петрофiзичних зв’язкiв, необхiдних для математичної постановки та розв’язання обернених задач двозондового нейтрон-нейтронного каротажу. http://www.nbuv.gov.ua/portal/all/reports/2009-02/09-02-18.pdf

bne: An algorithm of geophysical data inversion based on non-probabilistic presentation of a priori information and definition of Pareto-optimality Author Elena Kozlovskaya Affiliations Department of Geophysics, University of Oulu, POB 3000, FIN-90014, University of Oulu, Finland E-mail elena@babel.oulu.fi Journal Inverse Problems Create an alert RSS this journal Issue Volume 16, Number 3 Citation Elena Kozlovskaya 2000 Inverse Problems 16 839 doi: 10.1088/0266-5611/16/3/318 Article References Cited By Abstract This paper presents an inversion algorithm that can be used to solve a wide range of geophysical nonlinear inverse problems. The algorithm in based upon the principle of a direct search for the optimal solution in the parameter space. The main difference of the algorithm from existing techniques such as genetic algorithms and simulated annealing is that the optimum search is performed under control of a priori information formulated as a fuzzy set in the parameter space. In such a formulation the inverse problem becomes a multiobjective optimization problem with two objective functions, one of them is a membership function of the fuzzy set of feasible solutions, the other is the conditional probability density function of the observed data. The solution to such a problem is a set of Pareto optimal solutions that is constructed in the parameter space by a three-stage search procedure. The advantage of the proposed technique is that it provides the possibility of involving a wide range of non-probabilistic a priori information into the inversion procedure and can be applied to the solution of strongly nonlinear problems. It allows one to decrease the number of forward-problem calculations due to selective sampling of trial points from the parameter space. The properties of the algorithm are illustrated with an application to a local earthquake hypocentre location problem with synthetic and real data. Dates Issue 3 (June 2000) Received 5 Январь 2000 , in final form 24 Март 2000 Текст. к сожалению, мне недоступен

bne: Boris Mirkin, "Core Concepts in Data Analysis: Summarization, Correlation and Visualization" Springer | 2011 | ISBN: 0857292862 | 410 pages | Core Concepts in Data Analysis: Summarization, Correlation and Visualization provides in-depth descriptions of those data analysis approaches that either summarize data (principal component analysis and clustering, including hierarchical and network clustering) or correlate different aspects of data (decision trees, linear rules, neuron networks, and Bayes rule). Boris Mirkin takes an unconventional approach and introduces the concept of multivariate data summarization as a counterpart to conventional machine learning prediction schemes, utilizing techniques from statistics, data analysis, data mining, machine learning, computational intelligence, and information retrieval. Innovations following from his in-depth analysis of the models underlying summarization techniques are introduced, and applied to challenging issues such as the number of clusters, mixed scale data standardization, interpretation of the solutions, as well as relations between seemingly unrelated concepts: goodness-of-fit functions for classification trees and data standardization, spectral clustering and additive clustering, correlation and visualization of contingency data. The mathematical detail is encapsulated in the so-called “formulation” parts, whereas most material is delivered through “presentation” parts that explain the methods by applying them to small real-world data sets; concise “computation” parts inform of the algorithmic and coding issues. Four layers of active learning and self-study exercises are provided: worked examples, case studies, projects and questions.

bne: Спасут утопающих в данных К 2018 г. американские компании будут нуждаться только в 1,5 млн менеджеров (не считая других специалистов), обученных работать с большими объемами информации, выяснили консультанты McKinsey Global Institute. Сколько подобных работников понадобится в мировом масштабе, трудно даже представить Григорий Милов Ведомости 19.05.2011, 89 (2855) В 2010 г. мировые корпорации хранили более 7 экзабайт (миллионов терабайт) данных, а потребители — еще более 6 экзабайт. Для сравнения: суммарно это в 52 000 раз больше, чем хранится в библиотеке конгресса США, пишут консультанты McKinsey Global Institute в отчете Big Data: The next frontier for inovation, competition and productivity. Причем данные переполняют практически все секторы глобальной экономики. По оценкам консультантов, к 2009 г. каждая компания с числом сотрудников более 1000 человек хранила более 200 терабайт данных, что вдвое больше, чем содержало в 1999 г. хранилище данных Walmart. Кто спасет В большинстве отраслей, пишут консультанты, компании, умеющие работать с большими объемами данных, будут зарабатывать больше тех, кто с этой задачей справляется плохо. Но для того чтобы реализовать этот потенциал, понадобятся квалифицированные специалисты: к 2018 г. спрос на специалистов, способных работать с большими объемами данных, только в США может достичь 475 000 человек. Для сравнения: в 2008 г. в США работало 156 000 человек, получивших серьезную аналитическую подготовку. Такие специалисты — редкость, по оценкам консультантов, в 2010 г. американские университеты в сумме выпустили лишь около 25 000 профессионалов нужных специальностей. Консультанты выделили три типа специалистов: 1) эксперты в области анализа данных, которым нужна глубокая подготовка в области статистики, искусственного интеллекта, навык превращать огромные массивы данных в полезную для бизнеса информацию; 2) менеджеры и аналитики, выступающие в роли квалифицированных потребителей информации, извлеченной из массивов данных; 3) обслуживающий персонал, способный разрабатывать, внедрять, поддерживать аппаратное и программное обеспечение, необходимое для хранения и использования данных. Как ожидают консультанты, предложение специалистов во всех трех сегментах будет одним из ключевых ограничений на пути экономического роста. К 2018 г. дефицит высококвалифицированных специалистов по обработке данных может достичь 140 000-190 000 человек. Столь масштабный дефицит не устранится сам собой — придется заниматься переподготовкой сотрудников и, возможно, привлекать специалистов из-за рубежа. Кто разбогатеет Исследователи MGI выделили четыре сегмента экономики, в которых работа с данными может дать заметный эффект: здравоохранение, розница, производство и новая отрасль — обработка персональных геолокационных данных. В каждом из сегментов они оценили потенциальную выгоду от умелой работы с информацией. Только в американском здравоохранении потенциальная выгода может составить до $300 млрд, а в европейских правительственных организациях — 250 млрд евро добавленной стоимости в год. Американские розничные торговцы смогут увеличить чистую маржу на 60%, в производстве можно достичь сокращения капитальных инвестиций на 7%, а издержек на разработку новых продуктов — на 50%, утверждают консультанты. По оценкам McKinsey, к 2020 г. только провайдеры геолокационных услуг смогут заработать более $100 млрд, а компании, предоставляющие смежные услуги, — еще более $800 млрд. За чей счет За счет внутренних ресурсов проблему не решить даже в США. Статистика, полученная McKinsey, показывает, что число выпускников нужных специальностей растет в Китае (на 10,4% в год в период с 2004 по 2008 г.) и России (на 1,5% в год с 2004 по 2008 г.), но детальной информацией о качестве и месте работы, например, российских выпускников они не обладают. Поэтому проанализировать соотношение спроса и предложения на кадровых рынках развивающихся стран консультанты не смогли, говорит один из авторов исследования — Майкл Чуй. По его словам, уже сейчас специалистам, находящимся в одной стране, технически возможно обрабатывать данные, расположенные в другой. Но некоторые страны ограничивают права нерезидентов в области работы с определенными типами данных, отметил Чуй. Потребность в специалистах, обучающихся, например, у нас на кафедре, растет, говорит Владимир Мхитарян, заведующий кафедрой математической статистики Высшей школы экономики. Выпускников, по его словам, уже недостаточно. Многие студенты уже с третьего курса начинают работать в аналитических подразделениях государственных и частных компаний. Спрос на таких специалистов в России и дальше будет расти, причем нелинейно, прогнозирует Мхитарян. Потребность в профессионалах в области анализа и обработки данных растет очень быстро и это одна из причин, по которым «Яндекс» в 2007 г. открыл свою собственную Школу анализа данных, говорит Михаил Левин, менеджер академических программ «Яндекса». В 2007 г. набор состоял из 80 человек с небольшим конкурсом, а в 2010 г. — уже 120 человек с конкурсом до семи человек на место. В 2011 г. «Яндекс» планирует открыть отделения школы в нескольких городах, в том числе за границей. Статистика показывает, что примерно 50% выпускников школы приходят работать в «Яндекс». По словам Левина, выпускники вузов, даже ведущих, в большинстве случаев требуют дополнительной подготовки, чтобы начать работу в компании уровня «Яндекса». Компания нанимает людей и на открытом рынке труда, но в последнее время сталкивается с дефицитом специалистов. Бороться с ним «Яндекс» планирует как с помощью расширения собственной школы, так и благодаря сотрудничеству с ведущими техническими вузами страны (например, МФТИ, ВШЭ, МГУ), говорит Левин. Читайте далее:http://www.vedomosti.ru/newspaper/article/260439/spasut_utopayuschih_v_dannyh#ixzz1MmDy86qc

bne: ЧАШКОВ АНАТОЛИЙ ВАСИЛЬЕВИЧ МАТЕМАТИЧЕСКОЕ МОДЕЛИРОВАНИЕ ФИЛЬТРАЦИОННО-ЕМКОСТНЫХ СВОЙСТВ ГЕОЛОГИЧЕСКИХ СРЕД С ИСПОЛЬЗОВАНИЕМ ДАННЫХ ГЕОФИЗИЧЕСКИХ ИССЛЕДОВАНИЙ СКВАЖИН 25.00.10 – геофизика, геофизические методы поиска полезных ископаемых АВТОРЕФЕРАТ диссертации на соискание ученой степени кандидата физико-математических наук Видимо буду оппонентом Честно говоря далеко не все мне там нравится, но против замечаний автор не возражал Слишком много вопросов автор пытается охватить По моему с физ-мат там как-то IMHO с проблемами (разве задачи по трещинам) На геологию по Восточной Сибири ну явно маловато На технические также вряд-ли - софт не свой Тем не менее, по нынешним временам и это наверное неплохо И в целом отзыв планирую положительный

БНЕ_Gurg: В упрощенной постановке - выбор разбиения и одновременно оптимальных регрессоров в каждой из частей на которые разбиение идет Интересно в частности для восстановления акустики

bne: A History of Inverse Probability: From Thomas Bayes to Karl Pearson By Andrew I. Dale 1999 | 690 Pages | ISBN: 0387988076 This is a history of the use of Bayes theoremfrom its discovery by Thomas Bayes to the rise of the statistical competitors in the first part of the twentieth century. The book focuses particularly on the development of one of the fundamental aspects of Bayesian statistics, and in this new edition readers will find new sections on contributors to the theory. In addition, this edition includes amplified discussion of relevant work.

bne: Petroleum Science Volume 9, Number 3 (2012), 295-302, DOI: 10.1007/s12182-012-0212-y The sensitivity of the array resistivity log to mud filtrate invasion and its primary five-parameter inversion for improved oil water recognition Shaogui Deng, Qingtao Sun, Hu Li, Ningning Huo and Xuquan He In order to improve reservoir fluid recognition, the sensitivity of array resistivity response to the difference of the invasion properties in both oil-bearing layers and water layers is analyzed. Then the primary inversion is carried out based on the array resistivity log. The mud invasion process is numerically simulated based on the oil-water flow equation and water convection diffusion equation. The results show that the radial resistivity of a fresh mud-invaded oil-bearing layer presents complex distribution characteristics, such as nonlinear increase, increasing to decreasing and low resistivity annulus, and the resistive invasion profile of a water layer is monotonic. Under specific conditions, array resistivity log can reflect these changes and the array induction log is more sensitive. Nevertheless, due to the effect of factors like large invasion depth, reservoir physical and oil-bearing properties, the measured apparent resistivity may differ greatly from the actual mud filtrate invasion profile in an oil-bearing layer. We proposed a five-parameter formation model to simulate the complex resistivity distribution of fresh mud-invaded formation. Then, based on the principle of non-linear least squares, the measured array resistivity log is used for inversion with the Marquardt method. It is demonstrated that the inverted resistivity is typically non-monotonic in oil-bearing layers and is monotonic in water layers. Processing of some field data shows that this is helpful in achieving efficient reservoir fluid recognition.

bne: BAYESIANISM: SCIENCE OR PSEUDOSCIENCE? MARIO BUNGE Department of Philosophy, McGill University, Canada ABSTRACT This is a criticism of Bayesianism, the opinion that all probabilities are a matter of opinion, hence beyond objective tests. It is shown that the mathematical concept of a probability function makes no room for a person, and that in physics, chemistry and biology probabilities are objective quantities subject to calculation and measurement. It is also shown that the use of subjective probabilities in medicine and criminology is bound to lead to either nonsense or injustice. The upshot is that only the realistic interpretation of probability, as the quantitation of objective possibility, is legitimate. Jnte111ational Review ofVictimology. 2008, Vol. IS, pp.165-178 C A B Academic Publishers - Printed in Great Britain ================ Марио Бунге когда то известный аргентинский философ На русский переведены его книги про ПРИЧИННОСТЬ и ИНТУИЦИЮ

bne: МОГУТОВ АЛЕКСАНДР СЕРГЕЕВИЧ РАЗРАБОТКА МЕТОДОВ ПОВЫШЕНИЯ ДОСТОВЕРНОСТИ ПОДСЧЁТА ЗАПАСОВ УГЛЕВОДОРОДОВ НА ОСНОВЕ ТЕХНОЛОГИИ НЕЧЁТКОГО МОДЕЛИРОВАНИЯ Специальность 25.00.16 – Горнопромышленная и нефтегазопромысловая геология, геофизика, маркшейдерское дело и геометрия недр АВТОРЕФЕРАТ диссертации на соискание ученой степени кандидата технических наук Ухта 2012Работа выполнена в Ухтинском государственном техническом университете Научный руководитель: кандидат технических наук, Кулешов Владислав Евгеньевич Официальные оппоненты: Крапивский Евгений Исаакович доктор геолого-минералогических наук, профессор Санкт-Петербургский горный Университет, профессор кафедры транспорта и хранения нефти и газа Долгаль Александр Сергеевич доктор физико-математических наук, профессор Горный институт Уральского отделения РАН, ведущий научный сотрудник Ведущая организация: Филиал ООО «Газпром ВНИИГАЗ» в г. Ухта Защита состоится «25» мая 2012 г., в 10:00 на заседании диссертационного совета Д 212.291.01 в Ухтинском государственном техническом университете по адресу: 169300, Республика Коми, г. Ухта, ул. Первомайская, 13. С диссертацией можно ознакомиться в библиотеке УГТУ Автореферат разослан «24» апреля 2012 г. Учѐный секретарь диссертационного совета кандидат технических наук, профессор Н. М. Уляшева3 ОБЩАЯ ХАРАКТЕРИСТИКА РАБОТЫ Актуальность работы Одной из главных задач нефтегазопромысловой геологии является подсчѐт запасов углеводородов. Достоверность подсчѐта запасов определяется тем, на- сколько достоверно выполнено определение подсчѐтных параметров, таких как по- ристость и нефтенасыщенность. Существуют ряд подходов к определению пара- метров для подсчѐта запасов. При этом, часть значений этих параметров находится путѐм непосредственного изучения представительных образцов керна в лаборатор- ных условиях, а другая, большая часть, определяется с помощью промыслово- геофизических методов исследования. Если первые исследования дают вполне дос- товерные значения параметров, то промыслово-геофизические методы не позволя- ют непосредственно установить величины пористости и нефтенасыщенности, но оценивают величины геофизических параметров, которые связаны корреляцион- ными зависимостями с коллекторскими свойствами пород и характером их насы- щения. Более того, нередко возникают ситуации, когда при наличии одних и тех же данных в итоге получаются принципиально различные результаты. Для нахождения конкретных выражений этих зависимостей и параметров их характеризующих используются, в частности, приемы статистической обработки данных и, в конечном итоге, осуществляется замена реального экспериментального материала полученными законами и некоторой интегральной, т. е. общей для всей зависимости в целом, оценкой меры тесноты связи. В соответствии с принятой технологией далее найденный закон в виде уравнений связи между параметрами переносится на изучаемый объект (геологическую модель). Этот путь зачастую яв- ляется источником ошибочных заключений, которые могут привести от неверно выбранных условий вскрытия и режимов эксплуатации скважины до ошибок в точ- ках заложения скважин и геологических заключений о запасах и ресурсах углево- дородного сырья регионов. Поэтому автором предлагается рассматривать не от- дельные зависимости и взаимосвязи, а использовать данные в полном объѐме в ви- де нечѐтких множеств. Таким образом, необходима методика, позволяющая дать объективную оценку определения подсчѐтных параметров, которую возможно дифференциро- вать по достоверности. 4 Цель работы Разработка теории, методов и технологии, обеспечивающих повышение объ- ективности и надѐжности определения подсчѐтных параметров на основе техноло- гии нечѐткого вывода Мамдани. Основные задачи исследований 1. Адаптация принципов нечѐткого вывода Мамдани к задачам прогноза чи- словых параметров и развитие метода нечѐтких петрофизических композиций. 2. Разработка метода определения подсчѐтных параметров на основе метода нечѐтких петрофизических композиций. 3. Адаптация программного обеспечения, реализующего метод нечѐтких петрофизических композиций, для решения задач дифференциальной оценки под- счѐтных параметров. 4. Построение трѐхмерных геолого-геофизических моделей распределения подсчѐтных параметров месторождений углеводородов Тимано-Печорской нефте- газоносной провинции. 5. Анализ эффективности разработанной унифицированной схемы для оцен- ки достоверности и качества построенных геолого-геофизических моделей на при- мере месторождений углеводородов Тимано-Печорской нефтегазоносной провинции. http://www.ugtu.net/sites/default/files/avtoreferat_mogutov..pdf Нечаянно Google навел ;-)

viking23: повышение достоверности на основе не четкого моделирования. Главное задач нефтегазопромысловой геологии является поиск новых запасов:)) Достоверность определяется достоверностью. "Если первые исследования дают вполне достоверные значения параметров, то промыслово-геофизические методы не позволяют непосредственно установить величины пористости и нефтенасыщенности, но оценивают величины геофизических параметров, которые связаны корреляционными зависимостями с коллекторскими свойствами пород и характером их насыщения. " Человек любит слово достоверный. При этом эффект масштаба ему не знаком. Предлагаю ему изменить фразу на. Измерения Керна происходят на существенно меньшем масштабе, чем размер области с которой геофизической аппаратурой регистрируется сигнал. Иными словами методами скважиной геофизики усредняются фильтрационно-емкостные свойства среды. Неоднородностью среды и объясняется расхождение результатов измерений по ГИС и Керну. Сколько не удостоверяй параметры они могут расходиться, хоть засунь прибор в скважину, тем более что граничные условия на поверхности образца в лаборатории и в скважине различны:) Интересно какая это теория позволяет сделать повышенную объективность оценки параметров? Наверно теория безопасности рабочего места, по которой экспериментатору необходимо на работе не спать, дабы не испортить процесс измерения:)

B_N_E_8: Притом с высокой ДОСТОВЕРНОСТЬЮ

БНЕ_Home: Lorenz Biegler, George Biros, "Large-Scale Inverse Problems and Quantification of Uncertainty" 2011 | ISBN-10: 0470697431 | 388 pages This book focuses on computational methods for large-scale statistical inverse problems and provides an introduction to statistical Bayesian and frequentist methodologies. Recent research advances for approximation methods are discussed, along with Kalman filtering methods and optimization-based approaches to solving inverse problems. The aim is to cross-fertilize the perspectives of researchers in the areas of data assimilation, statistics, large-scale optimization, applied and computational mathematics, high performance computing, and cutting-edge applications. The solution to large-scale inverse problems critically depends on methods to reduce computational cost. Recent research approaches tackle this challenge in a variety of different ways. Many of the computational frameworks highlighted in this book build upon state-of-the-art methods for simulation of the forward problem, such as, fast Partial Differential Equation (PDE) solvers, reduced-order models and emulators of the forward problem, stochastic spectral approximations, and ensemble-based approximations, as well as exploiting the machinery for large-scale deterministic optimization through adjoint and other sensitivity analysis methods. Key Features: • Brings together the perspectives of researchers in areas of inverse problems and data assimilation. • Assesses the current state-of-the-art and identify needs and opportunities for future research. • Focuses on the computational methods used to analyze and simulate inverse problems. • Written by leading experts of inverse problems and uncertainty quantification. Graduate students and researchers working in statistics, mathematics and engineering will benefit from this book.

bne: НАУКА и ТЭК 2012 №4 стр.16-18 Б.Н. Еникеев (Главный петрофизик ЗАО "ПАНГЕЯ", b n e _ p a n g e a . ru) 40 лет оптимизационной петрофизической инверсии (горизонты и ловушки). 1. Халфин Л.А. Информационная теория интерпретации геофизических исследований ДАН СССР т.122 № 6 1958 2. Гольцман Ф.М. Cтатистические модели интерпретации Наука Ф/М. 1971 327cтр. 3. Еникеев Б.Н. Cистемный подход к статистической интерпретации геофизических данных в задачах с априорно известной структурой многомерных моделей. Тезисы семинара “Применение математических методов и ЭВМ в геологии” Алма-Ата 1974, стр. 85-87. (www.petrogloss.narod.ru) 4. Еникеев Б.Н., Кашик А.С., Чуринова И.М., Шпикалов Ю.А. Cистемный подход к оценке свойств пласта по данным каротажа (модели и методы). М.: ВНИИОЭНГ, 1980 (Обзорная информация, сер. Нефтегазовая геология и геофизика). 5. Meyer C., Subbit A. Global, А new Approach to Computer Processed Log Interpretation SPE 9346 55 Annual Fall Conference and Exhibition of the SPE 1980. 6. Cидорчук А.И., Кнеллер Л.Е., Гайфуллин Я.С. Использование идей оптимизации и идентификации при комплексной обработке данных каротажа. Математические методы идентификации моделей в геологии М.: МОИП 1983 стр.58-61. 7. Alberty Mark W., Hashmy Khaled H., Application of Ultra to Log Analysis, SPWLA 25th Annual Logging Symposium, 1984-Z 8. Еникеев Б., Чуринова И., Шпикалов Ю. Развитие методов количественной интерпретации ГИС в рамках АСОИГИС (проблемы и перспективы) Исследования и разработки в области нефтяной геофизики в странах-членах СЭВ т.2 М.: CЭВ 1988 cтр 33-40. 9. Peeters M., Visser R. A Comparison Of Petrophysical Evaluation Packages: Logic, Flame, Elan, Optima, and Ultra, The Log Analyst, Volume 32, Number 4, July - August, 1991 10. Элланский М.М., Еникеев Б.Н. Использование многомерных связей в нефтегазовой геологии М.: Недра 1991 208стр. 11. Зунделевич С.М., Еникеев Б.Н., Неяглова О.А. Некоторые принципиальные проблемы построения систем поддержки количественной интерпретации данных каротажа и пути их решения в оболочке PetroSoftShell //Проблемы интерпретации данных ГИС на ЭВМ. Сб.тр. Вып.2 Тюмень 1992 стр.19-25. 12. Quirein J. A., Truax J. A. and Perkins T., A Petrophysical Interpretation Framework Supporting the Development of Geological, Geophysical and Engineering Models, SPWLA 45th Annual Logging Symposium, 2004, Paper JJJ. 13. Mosegaard K, Tarantolla A. Probabilistic Approach to Inverse Problems In: International Handbook of Earthquake and Engineering Seismology, published by Academic Press for the International Association of Seismology and Physics of the Earth Interior, 2002. 14. Kazatchenko, E., Markov, M., Mousatov, A., Pervago, E., “Joint inversion of conventional well logs for evaluation of double-porosity carbonate formations”. Journal of Petroleum Science and Engineering, 2007, Vol. 56, No 4, pp. 252–266. 15. Carlos Torres-Verdín, Bovan K. George, Mojdeh Delshad, Richard Sigal, Farid Zouioueche, and Barbara Anderson Assessment of In-Situ Hydrocarbon Saturation in the Presence of Deep Invasion and Highly Saline Connate Water PETROPHYSICS, 2004 VOL. 45, NO. 2; P. 141–156 16. Enikeev B.N., Kazurov A.B. Development of technology of petrophysical knowledge synthesis and its application to formation evaluation. SPG-2010 Haiderabad [2010] 17. Еникеев Б.Н., Казуров А.Б.Опыт построения и использования алгоритмов петрофизической инверсии и настройки их параметров CD-ROM KazGeo [2010] 18. Еникеев Б.Н. Презентация возможностей ModERn в Лукойл (октябрь 2012) http://www.pangea.ru/ru/img_content/public/PetroExpert_ModERn_Enikeev.pdf 19. Greg Hodges VOODOO METHODS: DEALING WITH THE DARK SIDE OF GEOPHYSICS Fugro Airborne Surveys, Toronto ,Canada SAGEEP-2005 http://petrophysics.b.qip.ru/?1-0-0-00000030-001.001.001-0-0 20. Ritzer George The McDonaldization of Society 5. Pine Force Press 2008

bne: До этого были проблемы с ключом Сейчас впечатление что они сняты ModERn инсталлировали уфимцам под их ноутбук с Windows-7

bne: В Томском сборнике http://portal.tpu.ru/files/departments/publish/IPR_Erofeev_Isaev_geofiz_metod_.pdf на стр 177 - есть статья некого Михайлова В.Н. (Казань) на тему "О корректности используеиых методик интерпретации петрофизических данных" 1) Критикуется подгонка уравнений для проницаемости (дескать надо не логарифм, а напрямую) - автор не задумывается как распределены ошибки измерений 2) Критикуется подход расчета граничных по кумулянтам (но не исходный, с нормированием на 100%, а в форме приданной ему в Методическом руководстве) Оно бы и правильно, но прежде чем критиковать желательно четче ориентироваться в теме

bne: ОЦЕНКА ЭФФЕКТИВНОСТИ ПОВТОРНЫХ ГИДРАВЛИЧЕСКИХ РАЗРЫВОВ ПЛАСТА РАСПОЗНАВАНИЕМ ОБРАЗОВ в диссертации САБИТОВ РАЗИЛЬ РАЗИМОВИЧ ОБОСНОВАНИЕ И РАЗРАБОТКА МЕТОДА ОЦЕНКИ ЭФФЕКТИВНОСТИ ПОВТОРНЫХ ГИДРАВЛИЧЕСКИХ РАЗРЫВОВ ПЛАСТА Специальность 25.00.17 – Разработка и эксплуатация нефтяных и газовых месторождений Автореферат диссертации на соискание ученой степени кандидата технических наук Тюмень – 2012 Работа выполнена в Федеральном государственном бюджетном образовательном учреждении высшего профессионального образования «Тюменский государственный нефтегазовый университет» (ТюмГНГУ) Министерства образования и науки Российской Федерации Научный руководитель - кандидат технических наук Коротенко Валентин Алексеевич Официальные оппоненты: - Росляк Александр Тихонович, доктор технических наук, профессор, ФГБОУ ВПО «Национально исследовательский Томский политехнический университет», кафедра «ГРНМ», профессор; - Копытов Андрей Григорьевич, кандидат технических наук, ГП ХМАО - Югры «Научно - аналитический центр рационального недропользования им. В.И. Шпильмана», старший научный сотрудник. Ведущая организация - Филиал общества с ограниченной ответственностью «ЛУКОЙЛ - Инжиниринг» «КогалымНИПИнефть» в г. Тюмени Защита состоится 29 марта 2012 года в 14.00 часов на заседании диссертационного совета Д 212.273.01 при ТюмГНГУ по адресу: 625027, г. Тюмень, ул. 50 лет Октября, 38. С диссертацией можно ознакомиться в библиотечно-информационном центре ТюмГНГУ по адресу: 625027, г. Тюмень, ул. Мельникайте, 72 а, каб. 32. Автореферат разослан 29 февраля 2012 года. Ученый секретарь диссертационного совета, доктор технических наук, профессор Г.П. Зозуля ОБЩАЯ ХАРАКТЕРИСТИКА РАБОТЫ Актуальность проблемы Для нефтегазодобывающего комплекса России в настоящее время актуальны проблемы увеличения нефтеотдачи и вовлечения в разработку трудноизвлекаемых запасов нефти. На месторождениях Западной Сибири удельный вес залежей, приуроченых к низкопроницаемым и расчлененным коллекторам, составляет около 60 %. С целью вовлечения в разработку недренируемых запасов нефти применяется гидравлический разрыв продуктивного пласта (ГРП). По экспертным оценкам около трети запасов углеводородов можно извлечь только с использованием этой технологии. Так, например, на объекте ЮВ1 Урьевского месторождения дополнительная добыча нефти за счет ГРП составляет 80,9 % от текущих накопленных отборов нефти. Высокопроводящие трещины гидроразрыва позволяют увеличить дебит скважин в 2 – 3 раза и более. Также известно, что за период эксплуатации скважин после проведения ГРП значительно снижается проводимость трещин вследствие выноса проппанта и ее постепенного смыкания. Восстановление производительности этих скважин производят путем повторных ГРП, которые имеют неоднозначную результативность. На Нивагальском месторождении кратность увеличения дебитов жидкости после повторного ГРП в 1,5 раза выше по сравнению с первым, а на Урьевском месторождении ситуация диаметрально противоположная, так как подбор скважин-кандидатов для повторного ГРП осуществляется без обоснования его эффективности ввиду отсутствия соответствующих методик. Известные процедуры прогнозирования показателей повторного ГРП не позволяют учесть все значимые факторы, так как решения содержат много упрощающих предположений, сужающих круг применения полученных результатов и требуют больших затрат материальных и временных ресурсов. В этой связи разработка метода прогнозирования эффективности повторного ГРП и, следовательно, проектирования его показателей является в настоящее время актуальной задачей. Цель работы Повышение нефтеотдачи пластов на основе новой методики оценки эффективности повторного гидравлического разрыва пласта. Основные задачи исследования 1. Анализ методов прогнозирования эффективности гидравлического разрыва пласта. 2. Исследования параметров характеризующих эффективность ГРП и обоснование наиболее значимых для формирования базы образов для объектов ЮВ1 Нивагальского и Урьевского месторождений. 3. Разработка метода прогнозирования показателей эффективности повторного гидравлического разрыва пласта на основе теории распознавания образов. 4. Промысловая апробация разработанной методики на месторождениях ТПП «Лангепаснефтегаз». Объект и предмет исследования Объектом исследования является многократный процесс гидравлического разрыва продуктивного пласта; предметом – скважины эксплуатирующие низкопроницаемые объекты ЮВ1 Нивагальского и Урьевского месторождений. Научная новизна выполненной работы 1. Научно обосновано и практически подтверждена эффективность проведения повторного гидравлического разрыва в низкопроницаемых и расчлененных пластах. 2. Разработан метод прогнозирования показателей эффективности повторных ГРП с применением теории распознавания образов, учитывающей параметры пласта и технологические показатели первого ГРП. Практическая ценность и реализация Разработанный программный продукт для прогнозирования показателей эффективности повторного ГРП применяется в ТПП «Лангепаснефтегаз» на скважинах, эксплуатирующих низкопроницаемый расчлененный объект ЮВ1 Нивагальского и Урьевского месторождений в результате чего получено повышение накопленной добычи нефти на 15 %. Доказано, что разработанную методику прогнозирования показателей эффективности повторных ГРП можно использовать для прогнозирования других видов ГТМ с последующим выбором параметров, построения базы образов и показателей их эффективности. Основные защищаемые положения 1. Установлены 17 параметров, влияющих на эффективность повторных ГРП, из которых 4 характеризуют геолого-физические свойства пласта, 8 – технологические характеристики ГРП, 5 –эффективность первого ГРП. 2. Метод прогнозирования показателей эффективности повторных ГРП с применением теории распознавания образов, учитывающий геолого-физические параметры пласта, технологию проведения первого ГРП и параметры его эффективности. Соответствие диссертации паспорту научной специальности Область исследования соответствует паспорту специальности 25.00.17 – «Разработка и эксплуатация нефтяных и газовых месторождений», а именно: пункту 4 – «Технологии и технические средства добычи и подготовки скважинной продукции, диагностика оборудования и промысловых сооружений, обеспечивающих добычу, сбор и промысловую подготовку нефти и газа к транспорту, на базе разработки научных основ ресурсосбережения и комплексного использования пластовой энергии и компонентов осваиваемых минеральных ресурсов» и пункту 5 – «Научные основы компьютерных технологий проектирования, исследования, эксплуатации, контроля и управления природно-техногенными системами, формируемыми для извлечения углеводородов из недр или их хранения в недрах с целью эффективного использования методов и средств информационных технологий, включая имитационное моделирование геологических объектов, систем выработки запасов углеводородов и геолого-технологических процессов». Апробация результатов исследований Результаты диссертационной работы и ее основные положения докладывались и обсуждались на: Международном симпозиуме им. академика М.А. Усова «Проблемы геологии и освоения недр» (Томск, 2009 - 2011 гг.); Всероссийской научно – технической конференции «Новые технологии для ТЭК Западной Сибири» (Тюмень, 2010 гг.); Всероссийской научно – технической конференции «Геология и нефтегазоносность Западно-Сибирского мегабассейна» (Тюмень, 2011 гг.), Международной научно – технической конференции «Нефть и газ Западной Сибири» (Тюмень, 2011 г), Международном научном симпозиуме «Теория и практика применения методов увеличения нефтеотдачи пластов» (Москва, 2011 гг.), Научно-практической конференции «Проблемы нефтегазового комплекса Западной Сибири и пути повышения его эффективности» (Тюмень, 2011 гг.), семинарах кафедры «Разработка и эксплуатация нефтяных и газовых месторождений» ТюмГНГУ (Тюмень, 2009 – 2011 гг.). Публикации Результаты выполненных исследований отражены в 10 печатных работах, в том числе три работы в изданиях, рекомендованных ВАК РФ. Объем и структура работы Диссертационная работа изложена на 133 страницах машинописного текста, содержит 27 таблиц, 33 рисунка. Состоит из введения, четырех разделов, основных выводов и рекомендаций, списка использованных источников из 94 наименований. СОДЕРЖАНИЕ РАБОТЫ Во введении обоснована актуальность работы, сформулированы цель и задачи исследований, научная новизна и практическая значимость диссертационной работы. В первом разделе представлены результаты анализа методов прогнозирования и моделирования эффективности гидравлического разрыва пласта. Прогнозированию дебитов нефти и газа, которые могут быть получены при создании трещин различной длины и проводимости, посвящены работы многих отечественных и зарубежных исследователей. Большой вклад в решение этой проблемы внесли Г.И. Баренблатт, Ю.Е. Батурин, Ю.П. Желтов, А.Г. Загуренко, Р.Р. Ибатуллнн, Р.Д. Каневская, Р.Я. Кучумов, А.Г. Копытов, В.Д. Лысенко, И.Т. Мищенко, А.А. Поздняков, А.В. Стрекалов, И.Н. Стрижов, А.П. Телков, Р.Е. Теслюк, С.А. Христианович, Р.Т. Фазлыев, К.М. Федоров, Cinco-Ley Н, Gringarten А.С., Rarney H.J., и др. Разработаны программы, учитывающие гидроразрыв пласта при проектировании разработки, например, «ТехСхема» в ОАО «СургутНИПИнефть», «Delphor» в ОАО «ЛУКОЙЛ», «Hydra’Sym» в ТюмГНГУ. В работах Кучумова Р.Я. представлены результаты множества исследований по прогнозированию показателей эффективности ГРП на месторождениях ОАО «Славнефть - Мегионнефтегаз». Выявлены факторы, которые определяют эффективность технологии ГРП и могут быть использованы при подборе скважин для проведения ГРП и диагностировании эффективности этого метода. Им предложен метод потенциальных функций для подбора скважин и диагностирования эффективности проведения ГРП применение, которого обеспечило точность 75 %. Разработана методика для экспресс-оценки эффективности процесса ГРП, основанная на использовании результатов математического моделирования развития трещины ГРП и последующего влияния созданной трещины на изменение фильтрационных потоков в пласте. В работах Некрасова В.И., Глебова А.В., Ширгазина Р.Г., Вахрушева В.В., Апасова Р.Т. исследовались методы многомерного статистического анализа данных, применяемых при прогнозировании показателей эффективности ГРП: метод к-средних и канонических корреляций. Выявлено, что недостатком метода канонических корреляций является линейность, приводящая к значительным расхождениям исходных геолого-технологических параметров от своих средних значений и увеличению погрешности вычислений. В ООО «ЛУКОЙЛ-Западная Сибирь» разработан метод «data mining», основанный на определении набора исходных параметров, которые условно разбиты на следующие группы: геологическое строение и фильтрационно-емкостные свойства; продуктивность скважин; текущее энергетическое состояние пласта; состояние запасов; состояние системы разработки; технология ГРП. В соответствии со стандартной процедурой его применения технологическая цепочка решения задач прогнозирования включает в себя: визуальную инспекцию данных, отбраковку данных, заполнение пропусков; предварительный линейный анализ признакового пространства; обучение нелинейных систем прогнозирования; контроль качества обучения на тестовых выборках; прогноз (точность прогноза составляет 74 %). На основе результатов исследований, проведенных с использованием резистивных моделей в лабораторных условиях McGuire-Sikora, были выведены теоретические зависимости увеличения добычи после ГРП (как для газовых, так и для нефтяных скважин). Зависимости были разработаны при определенных допущениях, которые должны быть учтены при их использовании для оценки увеличения продуктивности, которые не отражают продуктивность скважины за длительный период. Этот момент очень важен для условий гидравлического разрыва пластов с низкой проницаемостью (< 1,0∙10-3 мкм2). Они основаны на однофазном течении несжимаемой жидкости. Это допущение делает их наиболее применимыми для нефтяных скважин. Использованная модель основывалась на размещении скважин с плотностью сетки скважин 16,2 га/скв. В результате анализа фактических операций ГРП на месторождениях ООО «ЛУКОЙЛ-Западная Сибирь» И.А. Виноградовой и Ю.И. Ивановой были определены значения критериальных параметров для 16 основных групп объектов разработки (по 3-6 группам пластов для каждого из четырех территориально-производственных предприятий), которое стало методическим руководством по подбору скважин-кандидатов для проведения ГРП. В ОАО «ЛУКОЙЛ» разработана методика прогноза эффективности ГРП в виде информационно-аналитической системы Delphor, состоящей в выявлении и интегрировании всех возможных зависимостей между параметрами, описывающими состояние скважины и ее окрестности до ГРП и его эффективностью, определяемой дебитом нефти и обводненностью. Зависимости устанавливаются и интегрируются с помощью искусственных нейронных сетей. При этом скважины, по которым они определяются, должны быть сравнимыми, то есть характеризоваться векторами параметров одной длины. Применение системы Delphor позволило повысить прогнозные показатели на 15-20 %. Недостатком является, не использование технологических параметров проведения ГРП Унифицированная задача отличается от обычного проектирования ГРП тем, что задается геометрия трещины, которая бы максимально увеличивала продуктивность скважины. С использованием двумерных моделей решается обратная задача проектирования. В зависимости от математических подходов различают полностью трехмерные модели, трехмерные модели с плоскими трещинами и псевдотрехмерные модели. Все программы имеют примерно одни и те же общие недостатки, связанные с несовершенством теории и методических подходов к решению задач проектирования и осуществления ГРП. Выявлено, что отсутствует метод расчета повторного ГРП при наличии ранее созданной сети трещин в прискважинной зоне. В работах В.О. Богопольского, Г.Х. Габитова, И.В. Генералова, Р.А. Ивакина, А.Ю. Кручинина, В.А. Коротенко, А.Х. Мирзаджанзаде, А.С. Макаряна, М.Ф. Пустовалова, М.Е. Рожкина, М.М. Рашидова, Р.Х. Санникова и др. рассматривалось применение теории распознавания образов в нефтегазопромысловой практике. Широкое применение теории распознавания образов в области бурения, эксплуатации скважин оборудованных УШГН, методов увеличения нефтеотдачи, определения типа залежи, оценки газоотдачи, проведенное А.Х. Миразаджанзаде, О.Л. Кузнецовым, К.С. Басниевым, З.С. Алиевым в нефтегазовой практике позволяет утверждать, что достоверность полученных результатов высокая. Поэтому применение теории распознавания образов обеспечит фундаментальную основу для разработки методики прогнозирования показателей эффективности повторного ГРП. ================================== Материал обучения - 36 строк Размерность - 17 параметров (включая гранулометрию и прочике свойства пропанта) Как-то IMHO не cильно соотносятся Надо бы там что-то дорисовывать в выборку

БНЕ_Home: http://www.complexsystems.net.au/content/software

bne: The Hybrid of Classification Tree and Extreme Learning Machine for Permeability Prediction in Oil Reservoir Chandra Prasetyo Utomo Abstract Permeability is an important parameter connected with oil reservoir. In the last two decades, artificial intelligence models have been used. The current best prediction model in permeability prediction is extreme learning machine (ELM). It produces fairly good results but a clear explanation of the model is hard to come by because it is so complex. The aim of this research is to propose a way out of this complexity through the design of a hybrid intelligent model. The model combines classification and regression. In order to handle the high range of the permeability value, a classification tree is utilized. ELM is used as a final predictor. Results demonstrate that this proposed model performs better when compared with support vector machines (SVM) and ELM in term of correlation coefficient. Moreover, the classification tree model potentially leads to better communication among petroleum engineers and has wider implications for oil reservoir management efficiency. Keywords: Permeability Prediction, Extreme Learning Machine, Classification Tree, Hybrid Intelligent Systems, Oil Reservoir, Regression Problem IJCSI International Journal of Computer Science Issues, Vol. 10, Issue 1, No 1, January 2013 ISSN (Print): 1694-0784 | ISSN (Online): 1694-0814

bne: Natural Resources Research, Vol. 10, No. 4, December 2001 (C °2001) A History of Regression and Related Model-Fitting in the Earth Sciences (1636?-2000) Richard J. Howarth1,2 Received 9 April 2001; accepted 24 September 2001 The (statistical) modeling of the behavior of a dependent variate as a function of one or more predictors provides examples of model-fitting which span the development of the earth sciences from the 17th Century to the present. The historical development of these methods and their subsequent application is reviewed. Bond’s predictions (c. 1636 and 1668) of change in the magnetic declination at London may be the earliest attempt to fit such models to geophysical data. Following publication of Newton’s theory of gravitation in 1726, analysis of data on the length of a 1 ± meridian arc, and the length of a pendulum beating seconds, as a function of sin 2 (latitude), was used to determine the ellipticity of the oblate spheroid defining the Figure of the Earth. The pioneering computational methods of Mayer in 1750, Boscovich in 1755, and Lambert in 1765, and the subsequent independent discoveries of the principle of ‘least squares’ by Gauss in 1799, Legendre in 1805, and Adrain in 1808, and its later substantiation on the basis of probability theory by Gauss in 1809 were all applied to the analysis of such geodetic and geophysical data. Notable later applications include: the geomagnetic survey of Ireland by Lloyd, Sabine, and Ross in 1836, Gauss’s model of the terrestrial magnetic field in 1838, and Airy’s 1845 analysis of the residuals from a fit to pendulum lengths, from which he recognized the anomalous character of measurements of gravitational force which had been made on islands. In the early 20th Century applications to geological topics proliferated, but the computational burden effectively held back applications of multivariate analysis. Following World War II, the arrival of digital computers in universities in the 1950s facilitated computation, and fitting linear or polynomial models as a function of geographic coordinates, ‘trend surface analysis,’ became popular during the 1950–60s. The inception of ‘geostatistics’ in France at this time by Matheron had its roots in meeting the evident need for improved estimators in spatial interpolation. Technical advances in regression analysis during the 1970s embraced the development of regression diagnostics and consequent attention to outliers; the recognition of problems caused by correlated predictors, and the subsequent introduction of ‘ridge regression’ to overcome them; and techniques for fitting errors-in-variables and mixture models. Improvements in computational power have enabled ever more computer-intensive methods to be applied. These include algorithms which are ‘robust’ in the presence of outliers, for example Rousseeuw’s 1984 Least Median Squares; nonparametric smoothing methods, such as kernel-functions, splines and Cleveland’s 1979 LOcally WEighted Scatterplot Smoother (LOWESS); and the Classification and Regression

БНЕ_Home: . Айзерман М. А., Браверман Э. М., Розоноэр Л. И. Метод потенциальных функций в теории обучения машин. М., «Наука», 1970. 2. Андерсон Т. Введение в многомерный статистический анализ. М., Физматгиз, 1963. 3. Багров А. Н., Вапник В. Н., Снитковский А. И. К прогнозу опасных явлений погоды. Метеорология и гидрология, № 3, 1974 4. Бонгард М. М., Проблема узнавания. М., «Наука», 1967. 5. Браиловский В.Л., Лунц А. Л., Формулировка задачи распознавания объектов со многими параметрами и методы ее решения. Известия АН СССР, Техническая кибернетика, № 1, 1969. 6. Браиловский В. Л., Лунц А. Л., Программа классификации многопараметрических объектов, основанная на отборе статистически полезных признаков. Сб. под ред. В. Н. Вапника «Алгоритмы обучения распознаванию образов». М., «Советское радио», 1973. 7. Брайнес С. Н., Биологическая и медицинская кибернетика. М., «Медицина», 1971. 8. Блэкуэлл Д., Гришик М. Л., Теория игр и статистических решений. М., ИЛ, 1958. 9. Вайнцвайг М. Н., Алгоритм обучения распознаванию образов «Кора». В сб. под ред. В. Н. Вапника «Алгоритмы обучения распознаванию .образов». М., «Советское радио», 1973. 10. Вайнцвайг М. Н., Об одном алгоритме распознавания двоичных кодов. Проблемы передачи информации, т. 2, вып. 3, 1966. 11. Вапник В. Н., Глазкова Т. Г., Червоненкис А. Я. Алгоритмы обучения распознаванию образов, использующие метод обобщенных портретов. Алгоритмы ОП-4, ОП-5, ОП-6, ОП-7. Сб. под ред. В. Н. Вапника «Алгоритмы обучения распознаванию образов». М., «Советское радио», 1973. 12. Baпник В. Н., Журавель А. А., Червоненкис А. Я., Алгоритмы обучения распознаванию образов, использующие метод обобщенных портретов. Алгоритмы ОП-1, ОП-2, ОП-3. Сб. под ред. В. Н. Вапника «Алгоритмы обучения распознаванию образов». М., «Советское радио», 1973. 13. Вапник В. Н., Лернер А. Я., Узнавание образов при помощи обобщенных портретов. Автоматика и телемеханика, т. 24, № 6, 1963. 14. Вапник В. Н., Лернер А. Я., Червоненкис А. Я. Системы обучения распознаванию образов при помощи обобщенных портретов. Известия АН СССР, Техническая кибернетика, № 1, 1965. 15. Вапник В. Н., Червоненкис А. Я., Об одном классе персептронов. Автоматика и телемеханика, т. 25, № 1, 1964. 16. Вапник В. Н., Червоненкис А. Я., Об одном классе алгоритмов обучения распознаванию образов. Автоматика и телемеханика, т. 25, № 6, 1964. 17. Вапник В. Н., Червоненкис А. Я., Алгоритмы с полной памятью и рекуррентные алгоритмы обучения распознаванию образов. Автоматика и телемеханика, т. 29, №4, 1968. 18. Вапник В. Н., Червоненкис А. Я., О равномерной сходимости частот появления событий к их вероятностям. Теория вероятностей и ее применения, т. XVI, № 2, 1971. 19. Вапник В. Н., Червоненкис А. Я., Теория равномерной сходимости частот появления событий к их вероятностям и задача поиска оптимального решения по эмпирическим данным. Автоматика и телемеханика, № 2, 1971. 20. Витушкин А. Г., Оценка сложности задачи табулирования. М®, Физматгиз, 1959. 21. Гнеденко Б. В. Курс теории вероятностей. М., Физматгиз, 1961. 22. Гладышев Е. Г., О стохастической аппроксимации. Теория вероятностей и ее применения, т. 10, № 2, 1965. 23. Глазкова Т. Г., Гурарий К. Н., Даниленко С. И., и др., Возможности применения ЭВМ для клинико-рентгенологической дифференциальной диагностики рака и доброкачественного поражения пищевода. Вестник радиологии и рентгенологии, № 2, 1971. 24. Головкин Б. А., Машинное распознавание и линейное программирование. М. «Советское радио», 1972. 25. Губерман Ш. А., Извекова М. Л., Xургин Я. И. Применение методов распознавания образов при интерпретации геофизических данных. Сб. «Самообучающиеся автоматические системы». М., «Наука», 1966. 26. Девятериков И. П., Пропой А. И., Цыпкин Я. З., О рекуррентных алгоритмах обучения распознаванию образов. Автоматика и телемеханика, № 1, 1967. 27. Ермольев Ю. М., О методе обобщенных стохастических градиентов о стохастических квазифейеровских последовательностях. Кибернетика, № 2, 1969. 28. Журавель А. А., Трошко Н. В., Эджубов Л. Г., Использование алгоритма обобщенного портрета для опознания образов в судебном почерковедении. В кн.: Правовая кибернетика. М., «Наука», 1970. 29. Журавлев Ю. И., Дмитриев А. Н., Кренделев Ф. Н., О математических принципах классификации предметов и явлений. Дискретный анализ. Сб. трудов ИМ СО АН СССР. Новосибирск, № 7, 1966. 30. Загоруйко Н. Г., Методы распознавания и их применения. М., «Советское радио»., 1972. 31. Ковалевский В. А. (редактор), Сб. «Читающие автоматы». Киев, «Наукова думка», 1965. 32. Ковалевский В. А., Современное состояние проблемы распознавания образов. Кибернетика, № 5, 1967. 33. Козинец Б. Н., Об одном алгоритме обучения линейного персептрона. Сб. «Вычислительная техника и вопросы программирования». ЛГУ, 1964. 34. Козинец Б. Н., Рекуррентный алгоритм разделения двух множеств. В сб. под ред. В. Н. Вапника «Алгоритмы обучения распознавания образов». М., «Советское радио», 1973. 35. Козинец Б. Н., Ланцман P. М., Соколов Б. М., Якубович В. А., Опознание и дифференциация почерков при помощи электронно-вычислительных машин. Сб. «Самообучающиеся автоматические системы». М., «Наука», 1966. 36. Колмогоров А. Н., Три подхода к определению понятия «количество информации». Проблемы передачи информации, т. 1, вып. 1, 1965. 37. Колмогоров А. Н., Фомин С. В., Элементы теории функций и функционального анализа. М., «Наука», 1968. 38. Кунин П. Е., Бояджан В. А., и др., Распознавание образов и дифференциальная диагностика. Вестник АМН СССР, N° 5, 1968. 39. Курилов Б. М., Получение достаточных характеристик при распознавании образов. Вычислительные системы. Сб. трудов ИМ СО АН СССР, вып. 22 Н, 1966. 40. Ле Кам Л., О некоторых асимптотических свойствах оценок максимума правдоподобия. Сб. Математика, М., 1960. 41. Логинов В. И., Хургин Я. И., Общий подход к проблеме распознавания образов. Сб. тр. МИНХ и ГП, «Недра», вып. 62. М., 1966. 42. Логинов Н. В., Метод стохастической аппроксимации. Автоматика и телемеханика, № 4, 1966. 43. Линник Ю. В., Метод наименьших квадратов. М., Физматгиз, 1962. 44. Литваков Б. М., О сходимости рекуррентных алгоритмов обучения распознаванию образов. Автоматика и телемеханика, № 1, 1968. 45. Лунц А. Л., Браиловский В. Л., Об оценке признаков, полученных в статистических процедурах распознавания. Известия АН СССР, Техническая кибернетика, № 3, 1969. 46. Минский М., Пейперт С., Персептроны. М., «Мир», 1971. 47. Морозов В. С., Морозова В. А., и др., Прогнозирование индивидуального срока службы электронных приборов с помощью метода обобщенных портретов. Электронная техника, серия 1. Электроника С. В. Ч., выпуск 9, 1969. 48. Неймарк Ю. И., Баталова З. С. и др. Распознавание образов и медицинская диагностика. М., «Наука», 1972. 49. Нильсон Н., Обучающиеся машины. М., «Мир», 1967. 50. Первозванский А. А., Распознавание абстрактных образов, как задача линейного программирования. Известия АН СССР, Техническая кибернетика, № 4, 1965. 51. Поляк Б. Т., Метод сопряженных градиентов в задачах на экстремум. Журнал вычислительной математики и математической физики, т. 9, № 4, 1969. 52. Поляк Б. Т., Цыпкин Я. З. Псевдоградиентные алгоритмы адаптации и обучения. Автоматика и телемеханика, № 1, 1973. 53. Пугачев В. С., Статистические проблемы теории распознавания образов. Тр. III всесоюзного совещания по автоматическому управлению. М., «Наука», 1967. 54. Розенблатт Ф., Принципы нейродинамики. Персептрон и теория механизмов мозга. М., «Мир», 1965. 55. Раудис Ш. Ю., Об определении объема обучающей выборки линейного классификатора. Сб. тр. ИМ СО АН СССР, вып. 28. Новосибирск, 1965. 56. Романов Л. Н., Дымникова Т. Р., Построение линейных разделяющих функций для классификации синоптических ситуации. В кн.: Статистические методы в метеорологии, ч. I, Новосибирск, 1969. 57. Рыжик И. М., Градштейн И. С., Таблицы интегралов, сумм, рядов и произведений. М., ГИТТЛ, 1951. 58. Себастиан Г. С., Процессы принятия решений при распознавании образов. «Техника», 1965. 59. Смирнов Н. В., Теория вероятностей и математическая статистика (Избранные труды). М., «Наука», 1970. 60. Тихонов А. Н., О регуляризации некорректно поставленных задач. ДАН СССР, 153, 1, 1963. 61. Тихонов А. Н., О решении некорректно поставленных задач и методе регуляризации. ДАН СССР, 181, 3, 1963. 62. Уилкс С., Математическая статистика. М., «Наука», 1967. 63. Урбах В. Ю., Дискриминантный анализ: основные идеи и приложения. Сб. Статистические методы классификации, вып. 1. МГУ, 1969. 64. Феллер В., Введение в теорию вероятностей и ее приложения. М., «Мир», 1964. 65. Француз А. Г., Некоторые вопросы статистической теории опознания образов. Сб. Бионика. М., «Наука», 1966. 66. Xинчин А. Я., Об основных теоремах теории информации. Успехи математических наук, т. XI, вып. 1 (67), 1956. 67. Цыпкин Я. З., Адаптация и обучение в автоматических системах. М., «Наука», 1968. 68. Цыпкин Я. З., Основы теории обучающихся систем. М., «Наука», 1970. 69. Ченцов Н. Н., Оценка неизвестной плотности распределения по наблюдениям. ДАН СССР, т. 147, № 1, 1962. 70. Якубович В. А., Некоторые общие теоретические принципы построения обучаемых опознающих систем. Сб. Вычислительная техника и вопросы программирования. ЛГУ, 1965. 71. Якубович В. А., Рекуррентные конечносходящиеся алгоритмы решения системы неравенств. ДАН СССР, т. 166, № 6, 1966. 72. Abramson N., Вrаverman D., Learning to Recognize Patters in a Random Environment. Trans. IKE IT-8, № 5, 1962. 73. Abramson N., Braverman D., Sebastian G., Pattern Recognition and Machine bearing. IEEE Trans, on Information Theory v. 1–9, № 4, 1963. 74. Andersоn T. W., Bahadur R. R., Classification into two multivariate normal distributions with different со variance matrics. The annals of Mathematical statistics, June., v. 133, № 2, 1962. 75. Daniel J. W., The conjugate gradient method for linear and nonlinear operator equation. SIAM Numer. Anal. 4, № 1, 1967. 76. Dvoretsky A., On stochastic approximation. Proceedings of the III Berkelly Symposium on Mathematical Statistics and Probability, v. 1, 1956. 77. Fisher R. A., Contributions to Mathematical Statistics. New-York, 1952 78. Fix I. E., Hodges J. L., Disiriminatory analysis; nonparametric discrimination: consistency properties. Report 4 of the USAF School of Aviation Medicine, Randolph Field, Texas, 1952. 79. G1ivenkо V. I., Sulla determination empirica di probabilita, Giornale dell’ Istitute Italiano degli Attuari, 4, 1933. 80. Hestenes M. R., Stiefe1 E., Method of conjugate gradient's for solving linear systems. J., Res. Nat. Bur. Standards, 49, № 6, 1952. 81. Highleyman W. H., Linear decision functions with applications to pattern recognition. Proc. IRE, № 6, 1962. 82. Kiefer E., Wolfowitz J., Stochastic estimation of the maximum of a regression function. Ann. Math. Stat. v. 123, № 3, 1952. 83. Ко1mоgоrоff A. N., Sulla determinatione empirica di una legge di distributione, Giornale dell' Istitute Italiano degli Attuari, 4, 1933. 84. Le Kam L., On some asymptotic properties of maximum like-hood estimates and related Bayes' estimates. Calif. Public. Statist. 11, 1953. 85. Le Kam L., On asymptotic theory of estimation and testing hypotheses. Proc. III Berkley Symposium on Math. Stat, and Probability, v. 1, 1956. 86. Mоtzkin Т., Sсhоenberg I., The relaxation method for linear inequalities. Canadian J. Math., 6, № 3, 1954. 87. Novikoff A., On convergence proofs for perceptrons. Proceedings of Symposium on Mathematical Theory of Automata. Polytechnic Institute of Brooklyn, v. XII, 1963. 88. Pагzen E., On consistent estimates of the spectrum of stationary time series. Ann. Math. Stat., vol. 28, 1957. 89. Robbins H., Monro S., A stochastic approximation method. Annals of Math. Stat., v. 22, № 1, 1951. 90. Rоsеnb1att F., The perseptron, a probability model for information storage and organization in the brain. Psychol. Rev., 65, 1958. 91. Selfridge O., Pattern recognition and learning. Proceeding of the Third London Symposium on Information Theory. New York, 1956. 92. Steinbuсh K., Piske U. A. W., Learning matrics and their application. IEEE Transaction of Electronic Computers, v. EC–12, № 6. Dec. 1963. 93. Utt1eу A. M., A theory of the mechanism of learning on the computation of conditional probabilities, Proc. 1 st. Int. Congress Cybernetics. Namur, 1956. 94. Widrоw В., Generalization and information storage in networks of adaline neurons. Self-Organizing Systems, 1962. Washington, 1962. 95. Widrоw В., «Bootsharp learning» in threshold logic. Proc. of the III Congress of the IFAC. London, 1966. 96. Wald A., Note on consistency of the most like-hood estimate. Ann. Math. Stat. 20, 1949.

bne: Google ннавел на диссертацию Филькиной (из Перми) https://www.google.ru/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&ved=0CCkQFjAA&url=http%3A%2F%2Fwww.dissercat.com%2Fcontent%2Fveroyatnostno-statisticheskoe-obosnovanie-fatsialnoi-zonalnosti-devonskikh-terrigennykh-otlo&ei=7cAEU-eDD-nZ4ATFr4DACQ&usg=AFQjCNGmgkcSC65D396kOzCPS8ZWLJmrpg&sig2=pnDs8PoonLIrQv-4VJkNvQ

БНЕ_Home: Statistical Well Log Analysis of the Bazhenov Formation Kulyapin P.S. and Sokolova T.F. Jason, a CGG Company, Vyatka Business Centre, 35, Vyatskaya str., bld. 4, Moscow, 127015, Russia. The statistical log analysis has been applied to study the multimineral composition and porosity of the Bazhenov unconventional play. Effective and secondary porosities, as well as net thicknesses, were evaluated using the well-log interpretation workflow designed at the Well Logging Department of Gubkin Russian State University of Oil and Gas. The advantages, limitations and drawbacks of the suggested approach are discussed and illustrated with case studies for several fields located in the Krasnoleninsky and Surgut arches of the West Siberian basin. Secondary porosity (fractured/vuggy zones) has been analysed from shear wave splitting effects. Additionally, the mechanical and elastic properties of the Bazhenov shale have been analysed using well logs, in order to identify the most brittle rocks in the formation, considered as the most favourable for fracking and completion. Key words: Bazhenov formation, source rocks, shale play, unconventionals, TOC, mineralogy, mineral volumes, well log analysis, petrophysics, statistical modelling, elastic moduli, rock mechanics, brittleness.

БНЕ_Home: Речь идет о применении технологий JASON Ссылаются в частности и на мою публикацию 2003 года на сей счет Инверсия идет по всему интервалу (без учета различия свойств компонент) Корреляции объемных содержаний не учитываются И, тем не менее, работа IMHO интересна и полезна Расширенная версия статьи принята в PETROPHYSICS и выйдет в планируемом EAGE сборнике

bne: В числе авторов знакомые мне Гладких и Петров Статья дошла через Google без атрибутов издания Я бы с удовольствием поспорил с авторами об оценках алгоритмов и роли нейронных сетей Может сложится в Питере или на Геомодели (но тогда придется лететь туда) Наверное проще завести семинар в ZOOM

bne: Халфин Л.А. Информационная теория интерпретации геофизических исследований ДАН СССР т.122 № 6 1958 Гольцман Ф.М. Cтатистические модели интерпретации Наука Ф/М. 1971 327cтр. Еникеев Б.Н. Cистемный подход к статистической интерпретации геофизических данных в задачах с априорно известной структурой многомерных моделей. Тезисы семинара “Применение математических методов и ЭВМ в геологии” Алма-Ата 1974, стр. 85-87. (www.petrogloss.narod.ru) Еникеев Б.Н., Кашик А.С., Чуринова И.М., Шпикалов Ю.А. Cистемный подход к оценке свойств пласта по данным каротажа (модели и методы). М.: ВНИИОЭНГ, 1980 (Обзорная информация, сер. Нефтегазовая геология и геофизика). Meyer C., Subbit A. Global, А new Approach to Computer Processed Log Interpretation SPE 9346 55 Annual Fall Conference and Exhibition of the SPE 1980. Cидорчук А.И., Кнеллер Л.Е., Гайфуллин Я.С. Использование идей оптимизации и идентификации при комплексной обработке данных каротажа. Математические методы идентификации моделей в геологии М.: МОИП 1983 стр.58-61. Alberty Mark W., Hashmy Khaled H., Application of Ultra to Log Analysis, SPWLA 25th Annual Logging Symposium, 1984-Z Еникеев Б., Чуринова И., Шпикалов Ю. Развитие методов количественной интерпретации ГИС в рамках АСОИГИС (проблемы и перспективы) Исследования и разработки в области нефтяной геофизики в странах-членах СЭВ т.2 М.: CЭВ 1988 cтр 33-40. Peeters M., Visser R. A Comparison Of Petrophysical Evaluation Packages: Logic, Flame, Elan, Optima, and Ultra, The Log Analyst, Volume 32, Number 4, July - August, 1991 Элланский М.М., Еникеев Б.Н. Использование многомерных связей в нефтегазовой геологии М.: Недра 1991 208стр. Зунделевич С.М., Еникеев Б.Н., Неяглова О.А. Некоторые принципиальные проблемы построения систем поддержки количественной интерпретации данных каротажа и пути их решения в оболочке PetroSoftShell //Проблемы интерпретации данных ГИС на ЭВМ. Сб.тр. Вып.2 Тюмень 1992 стр.19-25. Quirein J. A., Truax J. A. and Perkins T., A Petrophysical Interpretation Framework Supporting the Development of Geological, Geophysical and Engineering Models, SPWLA 45th Annual Logging Symposium, 2004, Paper JJJ. Mosegaard K, Tarantolla A. Probabilistic Approach to Inverse Problems In: International Handbook of Earthquake and Engineering Seismology, published by Academic Press for the International Association of Seismology and Physics of the Earth Interior, 2002. Enikeev, B.N., 2003, Fitting and Solution of Inverse Petrophysical Problem on the Basis of Combination of the Parametric and Nonparametric Interconnections [in Russian], Paper F176, presented at the 2003 EAGE Conference Geophysics of the 21st Century–The Leap into the Future. Kazatchenko, E., Markov, M., Mousatov, A., Pervago, E., “Joint inversion of conventional well logs for evaluation of double-porosity carbonate formations”. Journal of Petroleum Science and Engineering, 2007, Vol. 56, No 4, pp. 252–266. Еникеев Б.Н. О системном подходе к петрофизическому обоснованию комплексной количественной интерпретации каротажа Труды международной конференции: ТАТНЕФТЕГЕОФИЗИКЕ-60 лет, 219-229 Еникеев Б.Н. 40 лет оптимизационной петрофизической инверсии (горизонты и ловушки). НАУКА и ТЭК http://ru.scribd.com/doc/99411921/Enikeev-Tumen-2012, 16-18 Еникеев Б.Н., Флеркевич ИЯ ОПЫТ НАСТРОЙКИ ПРОГРАММ ОПТИМИЗАЦИОННОЙ ИНВЕРСИИ ПРИ МАЛОМ ОБЪЕМЕ АПРИОРНОЙ ИНФОРМАЦИИ. ГЕОМОДЕЛЬ-2012 http://EAGE.org J.A. Jarzyna, U. Woznicka, A. Cichy Computing Solutions for the Comprehensive Interpretation of Well Logs EAGE 69th Conference & Exhibition — London, UK, 11 - 14 June 2007 M. Dobroka, N. Szabo, T.Ormos, B. Kiss, J. Toth, I. Szabo Interval Inversion of Borehole Geophysical Data for Surveying Multimineral Rocks Near Surface 2008 – 14 th European Meeting of Environmental and Engineering Geophysics Kraków, Poland, 15 - 17 September 2008 L.J. Miranda, C. Torres-Verdin, J.Lucia Modeling Mud-filtrate Invasion Effects on Resistivity Logs to Estimate Permeability of Vuggy and Fractured Carbonate Formations (SPE-121136) 71st EAGE Conference & Exhibition — Amsterdam, The Netherlands, 8 - 11 June 2009 Carlos Torres-Verdín, Bovan K. George, Mojdeh Delshad, Richard Sigal, Farid Zouioueche, and Barbara Anderson Assessment of In-Situ Hydrocarbon Saturation in the Presence of Deep Invasion and Highly Saline Connate Water PETROPHYSICS, 2004 VOL. 45, NO. 2; P. 141–156 Pavel Kulyapin, Tatiana F. Sokolova A Case Study about Formation Evaluation and Rock Physics Modeling of the Bazhenov Shale PETROPHYSICS, VOL. 55, NO. 3 (JUNE 2014); PAGE 211–218; 9 FIGURES; 1 TABLE

bne: The SEISCOPE optimization toolbox: A large-scale nonlinear optimization library based on reverse communication Ludovic Mйtivier and Romain Brossier ABSTRACT The SEISCOPE optimization toolbox is a set of FORTRAN 90 routines, which implement first-order methods (steepestdescent and nonlinear conjugate gradient) and second-order methods (l-BFGS and truncated Newton), for the solution of large-scale nonlinear optimization problems. An efficient line-search strategy ensures the robustness of these implementations. The routines are proposed as black boxes easy to interface with any computational code, where such large-scale minimization problems have to be solved. Traveltime tomography, least-squares migration, or full-waveform inversion are examples of such problems in the context of geophysics. Integrating the toolbox for solving this class of problems presents two advantages. First, it helps to separate the routines depending on the physics of the problem from the ones related to the minimization itself, thanks to the reverse communication protocol. This enhances flexibility in code development and maintenance. Second, it allows us to switch easily between different optimization algorithms. In particular, it reduces the complexity related to the implementation of second-order methods. Because the latter benefit from faster convergence rates compared to first-order methods, significant improvements in terms of computational efforts can be expected.

bne: Примерно тоже самое можно было написать лет 20-40 назад Но приложить результаты на тестовых функциях Да и сейчас любознательно было ьы сравнить с генетическими алгоритмами ;-)

bne: Известия НАН РК. Серия геологии и технических наук. М 3. 2014 УДК 550.8.053 Г. Т. БОРИСЕНКО, А. К. ИСАГАЛИЕВА, А. А. ЕРМАГУЛОВА (Казахский национальный технический университет им. К. И. Сатпаева, г. Алматы) ФОРМИРОВАНИЕ ИНТЕРПРЕТАЦИОННОЙ МОДЕЛИ СЛОЖНОПОСТРОЕННЫХ ТЕРРИГЕННЫХ КОЛЛЕКТОРОВ ВОСТОЧНОГО БОРТА ПРИКАСПИЙСКОЙ ВПАДИНЫ Аннотация. На примере подсолевых терригенных отложений восточного борта Прикаспийской впадины с использованием многомерности петрофизических и геофизических связей рассмотрен способ формирования интерпретационной модели по определению емкостных свойств горных пород. При формировании интерпретационной модели использованы многомерные корреляционные связи при определении коэффициентов глинистости, пористости, диаграммы электрических методов, радиоактивного каротажа и акустического метода. Для построения системы интерпретационных моделей установлены связи между измеренными по скважинным параметрами с одной стороны, и расширенным набором характеристик продуктивных отложений с другой. Ключевые слова: коллектор; удельное электрическое сопротивление, интенсивности естественного гамма-излучения, радиационного гамма-излучения, интервальное время; модели, многомерность связей; пористость, проницаемость, глинистость. --- Приведена таблица данных каротажа и данных испытания Но отсчеты заданы интервально Забавно, но приведены ссылки и на Элланского и на К-К (2007)

БНЕ_Home: Alberty, M.W., and K.H. Hashmy (1984), Application of ULTRA to log analysis. In: 25th Annual Logging Symposium SPWLA, 10–13 June 1984, New Orleans, USA, Society of Petrophysicists andWell Log Analysts, Houston, paper 1984-Z, 1–17. Archie, G.E. (1942) The electrical resistivity log as an aid in determining some reservoir characteristics. Trans. Min. Metall. Petrol. Eng. 146: pp. 54-62 Baker, A. (1996) OPTIMA, eXpress reference manual, Baker Atlas. Western Atlas Int. Inc., Houston Ball, S.M., Chace, D.M., Fertl, W.H. (1987) The Well Data System (WDS): An advanced formation evaluation concept in a microcomputer environment. Proc. SPE Eastern Regional Meeting, 21–23 October 1987, Pittsburgh, USA. Society of Petroleum Engineers Inc., Richardson, pp. 61-85 Clavier, C., Coates, G., Dumanoir, J. (1984) Theoretical and experimental bases for the dual-water model for interpretation of shaly sands. SPE J. 24: pp. 153-168 DeWitte, L. (1955) A study of electric log interpretation methods in shaly formations. Trans. Am. Inst. Min. Met. Eng. 204: pp. 103-110 Dobrуka, M. (1988) On the absorption-dispersion characteristics of channel waves propagating in coal seams of varying thickness. Geophys. Prospect. 36: pp. 318-331 CrossRef Dobrуka, M., Szabу, N.P. (2001) The inversion of well log data using Simulated Annealing method. Publs. Univ. Miskolc Geosci. A Min. 59: pp. 115-137 Dobrуka, M., and N.P. Szabу (2002), The MSA inversion of openhole well log data. In: Intellectual Service for Oil and Gas Industry: Vol. 2. Analysis, Solutions, Perspectives, Publs. Ufa State Petroleum Technological University and University of Miskolc, 27–38. Dobrуka, M., Szabу, P.N. (2005) Combined global/linear inversion of welllogging data in layer-wise homogeneous and inhomogeneous media. Acta Geod. Geophys. Hung. 40: pp. 203-214 CrossRef Dobrуka, M., Szabу, N.P., Hursбn, L. (2003) The interpretation of well log data by means of Float Encoded Genetic Algorithm. EAGE Extended Abstracts, 2–5 June 2003, Stavanger, Norway. Eur. Assoc. Geoscientists and Engineers, Houten, pp. 1-4 Dobrуka, M., Kiss, B., Szabу, P.N., Tуth, J., Ormos, T. (2007) Determination of cementation exponent using an interval inversion method. EAGE Extended Abstracts, 11–14 June 2007, London, Great Britain. Eur. Assoc. Geoscientists and Engineers, Houten, pp. 1-4 Dobrуka, M., Szabу, P.N., Cardarelli, E., Vass, P. (2009) 2D inversion of borehole logging data for simultaneous determination of rock interfaces and petrophysical parameters. Acta Geod. Geophys. Hung. 44: pp. 459-479 CrossRef Gyulai, (1995) Investigating dipping bed structures by means of analytic forward modeling. Magyar Geofizika 36: pp. 40-67 Gyulai, , Ormos, T., Dobrуka, M. (2010) A quick 2-D geoelectric inversion method using series expansion. J. Appl. Geophys. 72: pp. 232-241 CrossRef Holland, J.H. (1975) Adaptation in Natural and Artificial Systems. Univ. Michigan Press, Ann Arbor Ingber, L. (1989) Very fast simulated re-annealing. Math. Comput. Model. 12: pp. 967-973 CrossRef Marquardt, D.W. (1959) Solution of nonlinear chemical engineering models. Chem. Eng. Prog. 55: pp. 65-70 Mayer, C., Sibbit, A. (1980) GLOBAL, a new approach to computer-processed log interpretation. Proc. 55th SPE Annual Technical Conference and Exhibition, 21–24 September 1980, Dallas, USA. American Institute of Mining, Mettalurgical, and Petroleum Engineers, Dallas, pp. 1-14 Menke, W. (1984) Geophysical Data Analysis: Discrete Inverse Theory. Academic Press, New York Metropolis, N., Rosenbluth, A.W., Rosenbluth, M.N., Teller, A.H., Teller, E. (1953) Equation of state calculations by fast computing machines. J. Chem. Phys. 21: pp. 1087-1092 CrossRef Michalewicz, Z. (1992) Genetic Algorithms + Data Structures = Evolution Programs. Springer-Verlag, Berlin Ormos, T., Daragу, A. (2005) Parallel inversion of refracted travel times of P and SH waves using a function approximation. Acta Geod. Geophys. Hung. 40: pp. 215-228 CrossRef Poupon, A., Leveaux, J. (1971) Evaluation of water saturation in shaly formations. The Log Analyst 12: pp. 3-8 Log Interpretation Principles/Applications. Schlumberger Educational Services, Houston Sen, M.K., Stoffa, P.L. (1995) Global Optimization Methods in Geophysical Inversion. Ser. Advances in Exploration Geophysics,^A.J. Berkhout. Elsevier, Amsterdam Serra, O. (1984) Fundamentals of Well-Log Interpretation: The Acquisition of Logging Data. Elsevier, Amsterdam Simandoux, P. (1963) Dielectric measurements in porous media and application to shaly formations. Rev. Inst. Fr. Pйtrol. 18: pp. 193-215 Szabу, N.P. (2004) Global inversion of well log data. Geophys. Trans. 44: pp. 313-329 Vozzoff, K., Jupp, D.L.B. (1975) Joint inversion of geophysical data. Geophys. J. Roy. Astr. Soc. 42: pp. 977-991 Waxman, M.H., Smits, L.J.M. (1968) Electrical conductivities in oil-bearing shaly sands. SPE J. 8: pp. 107-122

bne: Joint inversion of rock properties from sonic, resistivity and density well-log measurements Paolo Dell’Aversana, Giancarlo Bernasconi, Fabio Miotti and Diego Rovetta Geophysical Prospecting, 2011, 59, 1144–1154 doi: 10.1111/j.1365-2478.2011.00996.x ABSTRACT Well-log data are processed in order to derive subsurface physical parameters, namely rock porosity, fluid saturations and permeability. This step involves the selection and inversion of experimental constitutive equations, which are the link between the rock parameters and geophysical measurements. In this paper we investigate the rock parameter observability and the reliability of well-log data processing. We present a visual analysis of the constitutive equations and of the inverse problem conditioning, when using independently, or jointly, log data from different domains. The existence of a common set of rock properties (cross-properties) that influence different measurements, makes it possible to reduce the ambiguities of the interpretation. We select a test case in a reservoir scenario and we explore how to determine rock porosity and fluid saturation from sonic, conductivity and density logs. We propose a Bayesian joint inversion procedure, which is able to control the conditioning problems, to efficiently take into account input data and model uncertainty and to provide a confidence interval for the solution. The inversion procedure is validated on a real well-log data set.

bne: https://sites.google.com/site/polconf/#program https://sites.google.com/site/polconf/

bne: Interval inversion approach for an improved interpretation of well logs Mihбly Dobrуka1 , Norbert Pйter Szabу2 , Jуzsef Tуth3 , and Pйter Vass3 GEOPHYSICS, VOL. 81, NO. 2 (MARCH-APRIL 2016); P. D163–D175, 12 FIGS., 3 TABLES. ABSTRACT The quality analysis of well-logging inversion results has always been an important part of formation evaluation. The precise calculation of hydrocarbon reserves requires the most accurate possible estimation of porosity, water saturation, and shale and rock-matrix volumes. The local inversion method conventionally used to predict the above model parameters depth by depth represents a marginally overdetermined inverse problem, which is rather sensitive to the uncertainty of observed data and limited in estimation accuracy. To reduce the harmful effect of data noise on the estimated model, we have suggested the interval inversion method, in which an increase of the overdetermination ratio allows a more accurate solution of the well-logging inverse problem. The interval inversion method inverts the data set of a longer depth interval to predict the vertical distributions of petrophysical parameters in a joint inversion procedure. In formulating the forward problem, we have extended the validity of probe response functions to a greater depth interval assuming the petrophysical parameters are depth dependent, and then we expanded the model parameters into a series using the Legendre polynomials as basis functions for modeling inhomogeneous formations. We solved the inverse problem for a much smaller number of expansion coefficients than data to derive the petrophysical parameters in a stable overdetermined inversion procedure. The added advantage of the interval inversion method is that the layer thicknesses and suitably chosen zone parameters can be estimated automatically by the inversion procedure to refine the results of inverse and forward modeling. We have defined depth-dependent model covariance and correlation matrices to compare the quality of the local and interval inversion results. A detailed study using well logs measured from a Hungarian gas-bearing unconsolidated formation revealed that the greatly overdetermined interval inversion procedure can be effectively used in reducing the estimation errors in shaly sand formations, which may refine significantly the results of reserve calculation.

bne: Было бы наивно ожидать, что они начнуть всерьез изучать историю вопроса Достаточно увидеть,олностью игнорируют своего соотечественника Zoltan Barlai В итоге и их также забудут

bne: И с чего вдруг разлагать искомые свойства по полиномам Казалось бы вейвлеты естественней Похоже, что к практике эти ребята отношения не имеют

bne: Парето-оптимальные решения обратной задачи гравиметрии с неопределенной априорной информацией Кишман-Лаванова, Т.Н. Інші назви: Pareto-optimal solutions of the inverse problem of gravimetry with indeterminate a priori information URI: http://dspace.nbuv.gov.ua/handle/123456789/103727 Посилання: Парето-оптимальные решения обратной задачи гравиметрии с неопределенной априорной информацией / Т.Н. Кишман-Лаванова // Геофизический журнал. — 2015. — Т. 37, № 5. — С. 93-103. — Бібліогр.: 18 назв. — рос. Дата: 2015 Переглядів: 28 Завантажень: 13 Анотація: Розглянуто теоретичні аспекти розв'язування оберненої нелінійної задачі гравіметрії в умовах невизначеності апріорної інформації. Апріорну інформацію описано за допомогою нечітких множин. Одноцільову геофізичну задачу з невизначеною апріорною інформацією трансформовано в багатокритеріальну задачу оптимізації. Одним із критеріїв є функція належності нечіткої множини можливих розв'язків. Розв'язком задачі є множина Парето-оптимальних розв'язків, яку конструйовано в параметричному просторі за допомогою триетапного алгоритму пошуку. Перевагою запропонованого методу є те, що він забезпечує можливість включення широкого інтервалу неймовірнісної апріорної інформації до процедури інверсії та може бути застосований для розв'язування сильно нелінійних задач. Це дає змогу зменшити кількість прямих обчислень задачі вибірковим моделюванням пробних точок у параметричному просторі. Наведено тестовий приклад роботи алгоритму в застосуванні до оберненої задачі гравіметрії для однієї контактної поверхні. The paper discusses theoretical aspects of solving the nonlinear inverse problem of gravimetry with uncertainty of a priori information. The a priori information is described by fuzzy sets. Special-purpose geophysical problem with uncertain a priori information is transformed into a multi-objective optimization problem. One of the criteria is the membership function of a fuzzy set of possible solutions. Solution of the problem is a set of Pareto-optimal solutions, which is constructed in the parametric space applying a three-step search algorithm. The advantage of the proposed method is that it provides a possibility of including the wide range of non- probabilistic a priori information in the inversion procedure and can be applied to the solution of highly nonlinear problems. This reduces the number of direct computing problems by selective modeling of sample points in the parametric space. A test example has been given of the algorithm applied to the inverse problem of gravimetry for a single contact surface.

bne: Получил любопытную статью от С.А. Корчагина из Самары, (коллега В.Чемоданова) УДК 519.25 СОВМЕСТНОЕ ИСПОЛЬЗОВАНИЕ ВЗАИМНО ОБРАТНЫХ ФУНКЦИЙ ПРИ РЕГРЕССИОННОМ АНАЛИЗЕ И РЕШЕНИИ ОБРАТНЫХ ЗАДАЧ Е.В. Гусева, С.А. Корчагин Автоматизация, телемеханизация и связь в нефтяной промышленности 6/2017 У меня немного иной взгляд и представление об алгоритме обработки (о чём ему и написал)

bne: Uncertainty reduction through geologically conditioned petrophysical constraints in joint inversion Jérémie Giraud, Evren Pakyuz-Charrier, Mark Jessell, Mark Lindsay, Roland Martin and Vitaliy Ogarko Right Running Head: Conditioned petrophysical constraints Geophysics (2017) 82 (6): ID19-ID34. The integration of probabilistic geological modeling permits more accurate retrieval of model geometry, and better constrains the solution while still satisfying the statistics derived from geological data. The analysis of statistical indicators at each step of the workflow shows that 1) the inversion methodology is effective when applied to complex geology, and 2) the integration of prior information and constraints from geology and petrophysics significantly improves the inversion results while decreasing uncertainty. Lastly, the analysis of uncertainty to the integration of the conditioned petrophysical constraints also shows that, for this example, the best results are obtained for joint inversion using petrophysical constraints spatially conditioned by geological modeling.

BNE_HOME: Uncertainty reduction in joint inversion using geologically conditioned petrophysical constraints : Giraud, Jérémie; University of Western Australia, Centre for Exploration Targeting Pakyuz-Charrier, Evren; University of Western Australia, Centre for Exploration Targeting Jessell, Mark; University of Western Australia, Centre for Exploration Targeting Lindsay, Mark; University of Western Australia, Centre for Exploration Targeting Martin, Roland; Observatoire Midi-Pyrenees Ogarko, Vitaliy; University of Western Australia, Centre for Exploration Targeting ABSTRACT 14 We introduce a joint geophysical inversion workflow that aims to improve subsurface 15 imaging and decrease uncertainty by integrating petrophysical constraints and geological 16 data. In this framework, probabilistic geological modeling is used as a source of information 17 to condition the petrophysical constraints spatially and to derive starting models. The 18 workflow then utilizes petrophysical measurements to constrain the values retrieved by 19 geophysical joint inversion. The different sources of constraints are integrated into a least20 square framework to capture and integrate information related to geophysical, petrophysical 21 and geological data. This allows us to quantify the posterior state of knowledge and to 22 calculate posterior statistical indicators. To test this workflow, using geological field data we 23 have generated a set of geological models, which we used to derive a probabilistic geological 24 model. In this synthetic case study, we show that the integration of geological information 25 and petrophysical constraints in geophysical joint inver-sion can reduce uncertainty and 26 improve imaging. In particular, the use of petrophysical constraints retrieves sharper 27 boundaries and better reproduces the statistics of the observed petrophysical measurements. Page 1 of 60 GEOPHYSICS 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 This paper presented here as accepted for publication in Geophysics prior to copyediting and composition. © 2017 Society of Exploration Geophysicists. Downloaded 08/04/17 to 132.203.227.63. Redistribution subject to SEG license or copyright; see Terms of Use at http://library.seg.org/ 2 28 The integration of probabilistic geological modeling permits more accurate retrieval of model 29 geometry, and better constrains the solution while still satisfying the statistics derived from 30 geological data. The analysis of statistical indicators at each step of the workflow shows that 31 1) the inversion methodology is effective when applied to complex geology, and 2) the 32 integration of prior information and constraints from geology and petrophysics significantly 33 improves the inversion results while decreasing uncertainty. Lastly, the analysis of 34 uncertainty to the integration of the conditioned petrophysical constraints also shows that, for 35 this example, the best results are obtained for joint inversion using petrophysical constraints 36 spatially conditioned by geological modeling.

bne: MinInversion: A Program for Petrophysical Composition Analysis of Geophysical Well Log Data Adewale Amosu * and Yuefeng Sun Abstract: Knowledge of the composition (mineral and fluid proportions) of rock formation lithologies is important for petrophysical and rock physics analysis. The mineralogy of a rock formation can be estimated by solving a system of linear equations that relate a class of geophysical log measurements to the petrophysical properties of known minerals and fluids. This method is useful for carbonate rocks with complex mineralogies and a wide range of other lithologies. Although this method of linear inversion for rock composition is well known, there are no interactive, open-source programs for routinely estimating rock mineralogy from standard digital geophysical wireline logs. We present an interactive open-source program, MinInversion, for constructing a balanced system of linear equations from digital geophysical logs and estimating the rock mineralogy as an inverse problem. MinInversion makes use of a library of petrophysical properties that can be easily expanded and modified by the users. MinInversion also provides several options for solving the system of linear equations and executing the linear matrix inversion including least squares, LU-decomposition and Moore-Penrose generalized inverse methods. In addition, MinInversion enables the estimation of the joint probability distributions for constituent minerals and measured porosity. The joint probability distributions are useful for revealing and analyzing depositional or diagenetic composition trends that affect porosity. The program introduces ease and flexibility to the problems of rock formation composition analysis and the study of the effects of rock composition on porosity.

bne: Component DENSITY (g/cc) DT (sec/ft) DTS (sec/ft) PE Quartz 2.65 55.5 88.8 1.82 Shale 2.6 62.5 150 3.42 Calcite 2.71 47.2 89.9 5.09 Clays 2.65 64.3 98.9 3.03 Dolomite 2.87 43.9 74.8 3.13 Anhydrite 2.95 50 85 5.08 Gypsum 2.35 52.4 85.4 4.04 Muscovite 2.83 47.2 91.1 2.4 Biotite 3.2 55.5 100.6 8.59 Kaolinite 2.64 64.3 101.7 1.47 Glauconite 2.83 55.5 157.4 4.77 Illite 2.77 64.3 98.9 3.03 Chlorite 2.75 55.5 61.3 4.77 Orthoclase 2.54 68.9 84.9 2.87 Siderite 3.91 43.9 84.9 14.3 Pyrite 5 39.6 55.9 16.4 Halite 2.03 6.7 114.5 4 FreshWater 1 205 쀀999.25 0.36 Brine 1.1 188 쀀999.25 0.81 Oil 0.8 238 쀀999.25 0.12

bne: ФЕДОР МАРКОВИЧ ГОЛЬЦМАН Федор Маркович Гольцман родился в 1924 г. в Ленинграде. Отец М.И. Гольцман — профессор Арктического института (геофизика атмосферы); мать М.Ф. Шоллар —артистка Мариинского театра. В 1942 г. Федор Маркович окончил с отличием среднюю школу. С 1942 по 1945 г. работал механиком в ОКБ военного завода №224 министерства авиастроения, пройдя путь от ученика до механика 6-го разряда. В 1945 г. Федор Маркович поступил на второй курс физического факультета Ленинградского государственного университета, который окончил в 1949 г. по специальности «радиофизика». С 1949 г. работал научным сотрудником, ассистентом, старшим научным сотрудником, доцентом, заведующим лабораторией, профессором физического факультета и Физического института Санкт-Петербургского университета. Кандидат (1952 г.), доктор (1962 г.) физико-математических наук. Автор семи монографий и более 180 научных статей в области информационно-статистических теории и методов интерпретации геофизических данных. С 1965 г. читал общие курсы лекций по физике и по статистическим методам интерпретации наблюдений на физическом и геологическом факультетах Ленинградского (С.-Петербургского) университета. Под руководством Ф.М. Гольцмана защитили кандидатские диссертации 23 специалиста, многие из которых стали докторами наук. С 1956 по 1985 годы Федор Маркович был руководителем и ответственным исполнителем более 40 хоздоговоров с ВНИИГеофизикой, ВИРГом и многими другими производственными геофизическими организациями. Участвовал в научно-исследовательских геофизических экспедициях ЛГУ, ВНИИГеофизики, Восточного, Краснодарского, Казахского, Узбекского геофизических трестов, Мингечаургэсстроя, Красноярской ГЭС, морской экспедиции АЗНИИ по добычи нефти. Многократно выезжал в загранкомандировки для чтения лекций и научной работы, участвовал во многих международных и российских конференций. Научную деятельность Ф.М. Гольцман начал в 1948 г. будучи студентом 5-го курса физического факультета ЛГУ, дипломной работой на тему «Переходные процессы в приемно-усилительных схемах» (руководитель проф. Г.И. Петрашень). Работа была опубликована в журнале «Радиотехника». Г.И. Петрашень, ученик известного математика акад. В.И. Смирнова, привил Ф.М. Гольцману склонность к теоретической работе в области прикладной математической физики, определив характер его дальнейшей деятельности. После окончания университета в 1949 г. Федор Маркович работал по распределению в лаборатории радиоспектроскопии Физического института Ленинградского университета, где используя навыки механика, полученные во время работы на военном заводе, сконструировал и построил установки для изучения в сантиметровом диапазоне радиоволн свойств молекул твердых и жидких тел. Результаты этих работ были опубликованы в центральных физических журналах и легли в основу кандидатской диссертации, защищенной в университете в 1952 г. После защиты Ф.М. Гольцман был лишен допуска к секретной работе и переведен на кафедру геофизики физического факультета, где занимался разработкой многоканальной сейсморазведочной аппаратуры для экспедиций кафедры на строительствах Мингечаурской и Красноярской ГЭС. В 1955 г. в Физическом институте ЛГУ была создана лаборатория динамики упругих сред, занимавшаяся под руководством проф.Г.И. Петрашеня вопросами теории и практики распространения сейсмических волн. Ф.М. Гольцман был переведен в эту лабораторию на должность старшего научного сотрудника. Тематика его занятий сначала сводилась к аппаратурному обеспечению экспедиций лаборатории, но в 1956 г. он заинтересовался частотным анализом и синтезом, а также дискретными линейными преобразованиями сейсмических сигналов. Вокруг него сформировалась группа учеников и последователей из числа студентов старших курсов и сотрудников прикладных организаций, положившая начало его будущей школе. В те годы развитие электронной вычислительной техники еще находилось в начальной стадии, поэтому весьма актуальными были разработанные и построенные этой группой электромеханические счетнорешающие устройства для частотного анализа и синтеза сигналов, а также приборы для дискретных линейных преобразований сейсмических импульсов и аномальных потенциальных полей. Эти приборы были выпущены малой серией заводом «Геологоразведка». Идеи, положенного в основу частотного анализа-синтеза впоследствии были повторены за границей в программах для ЭВМ быстрого преобразования Фурье (Кули-Тюки). Результаты работ группы Ф.М. Гольцмана по исследованию свойств сигналов и помех на этих приборах легли в основу кандидатских диссертаций ряда сотрудников группы. В 1956 г. началось новое направление работ Ф.М. Гольцмана и его учеников в области информационно-статистической теории и методов интерпретации геофизических данных. Основанием к появлению этого направления послужила статья Л.А. Халфина «Информационная теория интерпретации геофизических исследований», опубликованная в «Докладах АН СССР» в 1955 г., а также книга Ф.М. Вудворта «Теория вероятностей и теория информации с применениями в радиолокации», изданная в 1955 г. На начальной стадии (1955—1962 гг.) работы Ф.М. Гольцмана в этой области были сильно привязаны к его предыдущей тематики по Фурье-анализу и дискретным линейным преобразованиям сигналов, но уже с учетом влияния случайных помех. В этот период им были разработаны частотные методы и методы группирования сигналов, теоретически обоснован метод регулируемого направленного приема (РНП), предложены теория и методы фильтрации скоростей распространения сейсмических волн. В дальнейшем теория дискретных линейных преобразований была распространена на случай интерпретации аномальных потенциальных полей. Группа последователей и учеников Ф.М. Гольцмана существенно разрослась, в ней стали участвовать сотрудники многих производственных организаций и прикладных институтов на хоздоговорных началах, а также аспиранты и соискатели. Работы этого периода легли в основу докторской диссертации Ф.М. Гольцмана на тему «Основы интерференционного приема сейсмических волн», защищенной в ИФЗ СССР в 1961 г., а также ряда кандидатских диссертаций его учеников. В 1964 г. опубликована монография «Основы теории интерференционного приема регулярных волн». С 1962 г. в работах Ф.М. Гольцмана наметился окончательный переход в область использования результатов теории вероятностей, математической статистики и теории информации при интерпретации любых наблюдений. Большое значение имела серия обсуждений работ группы Ф.М. Гольцмана на семинарах лаборатории теории вероятностей и математической статистики академика Ю.В. Линника (ЛОМИ АН СССР), а также возможность использования ЭВМ, а в дальнейшем и персональных компьютеров. В 1965 г. Ф.М. Гольцман утвержден в звании профессора по кафедре общей физики физического факультета ЛГУ. С этого времени он читает общие курсы лекций по физике, методам решений обратных физических задач, а также по основам метрологии. В этот период школа профессора Ф.М. Гольцмана существенно разрослась. В нее вошли аспиранты и сотрудники разных кафедр физического факультета ЛГУ, представители прикладных институтов, сотрудники ряда производственных геофизических трестов, организации ГДР, Венгрии, ЧССР, КНР. Поступали неоднократные предложения участвовать в научных конференциях и сотрудничать с геофизическими организациями западных стран (США, ФРГ, Швейцария, Израиль и др.). Эти предложения, однако, не получили никакого развития, так как профессор Ф.М. Гольцман считался невыездным в капиталистические страны. В результате, его труды и труды его учеников не получили должного признания, а многие результаты статистической теории интерпретации были впоследствии «вторично открыты» в трудах западных ученых (B.L.N. Kennett, A. Tarantola и др.) В конце 60-х годов начала оформляться единая логически увязанная концепция информационно-статистических теории и методов наблюдений. Теоретические основы были обобщены в книге Ф.М. Гольцмана «Статистические модели интерпретации» (1971 г.), результаты работ его учеников в разных областях геофизики вошли в сборник «Статистические методы интерпретации геофизических данных» (1971 г.). После 80-х годов внедрению статистической теории и методов способствовало появление персональных компьютеров. Были разработаны алгоритмы разделения волн и построения разрезов в сейсмике, решались обратные задачи по данным грави-магнито-электроразведки, уточнялись постановки обратных задач с учетом влияния мешающих параметров с применением методов статистического моделирования. Большое влияние уделялось оценкам эффективности решений. Были построены оптимальные статистические алгоритмы и разработаны компьютерные технологии для решения прогнозных и классификационных задач (INTERACT, MULTALT, COMPLEX и др.). Полученные результаты изложены в монографиях «Физический эксперимент и статистические выводы» (1982) и «Статистическая интерпретация магнитных и гравитационных аномалий» (1983). Ф.М. Гольцман до конца своих дней был полон новых идей, оригинальных предложений, многие из которых реализовывал сам, разрабатывая компьютерные программы. Он был очень талантливым человеком, исключительно порядочным по отношению к своим коллегам, искренне любящим науку и всячески способствовавшим ее развитию. Умер он в 1999 г., внезапно, направляясь в университет принимать экзамены у студентов... Развитое Ф.М. Гольцманом информационно-статистическое направление в наши дни продолжает активно совершенствоваться, находя все большее число сторонников, что свидетельствует о его исключительной перспективности.

bne: When Petrophysics Meets Big Data: What can Machine Do? Chicheng Xu, Aramco Services Company: Aramco Research Center; Siddharth Misra, The University of Oklahoma; Poorna Srinivasan, Aramco Services Company: Aramco Research Center; Shouxiang Ma, Saudi Aramco Copyright 2019, Society of Petroleum Engineers This paper was prepared for presentation at the SPE Middle East Oil and Gas Show and Conference held in Manama, Bahrain, 18-21 March 2019. This paper was selected for presentation by an SPE program committee following review of information contained in an abstract submitted by the author(s). Abstract Petrophysics is a pivotal discipline that bridges engineering and geosciences for reservoir characterization and development. New sensor technologies have enabled real-time streaming of large-volume, multi-scale, and high-dimensional petrophysical data into our databases. Petrophysical data types are extremely diverse, and include numeric curves, arrays, waveforms, images, maps, 3-D volumes, and texts. All data can be indexed with depth (continuous or discrete) or time. Petrophysical data exhibits all the "7V" characteristics of big data, i.e., volume, velocity, variety, variability, veracity, visualization, and value. This paper will give an overview of both theories and applications of machine learning methods as applicable to petrophysical big data analysis. Recent publications indicate that petrophysical data-driven analytics (PDDA) has been emerging as an active sub-discipline of petrophysics. Field examples from the petrophysics literature will be used to illustrate the advantages of machine learning in the following technical areas: (1) Geological facies classification or petrophysical rock typing; (2) Seismic rock properties or rock physics modeling; (3) Petrophysical/geochemical/geomechanical properties prediction; (3) Fast physical modeling of logging tools; (4) Well and reservoir surveillance; (6) Automated data quality control; (7) Pseudo data generation; and (8) Logging or coring operation guidance. The paper will also review the major challenges that need to be overcome before the potentially game-changing value of machine learning for petrophysics discipline can be realized. First, a robust theoretical foundation to support the application of machine leaning to petrophysical interpretation should be established; second, the utility of existing machine learning algorithms must be evaluated and tested in different petrophysical tasks with different data scenarios; third, procedures to control the quality of data used in machine leaning algorithms need to be implemented and the associated uncertainties need to be appropriately addressed. The paper will outlook the future opportunities of enabling advanced data analytics to solve challenging oilfield problems in the era of the 4th industrial revolution (IR4.0).

bne: Deep neural network application for 4D seismic inversion to changes in pressure and saturation: Optimizing the use of synthetic training datasets Gustavo Côrte1∗, Jesper Dramsch2, Hamed Amini1,† and Colin MacBeth1 1Heriot-Watt University, Edinburgh, EH14 4AS, UK, and 2Technical University of Denmark, Copenhagen, Denmark Geophysical Prospecting, 2020, 68, 2164–2185 В данной работе мы решаем задачу количественной оценки динамики коллектора. изменения свойств в течение периода производства, непосредственно из четырехмерного сейсмические данные в амплитудной области. Мы используем глубокую нейронную сеть для инвертирования четырехмерные карты амплитуд сейсмических волн для одновременных изменений давления, водо- и газонасыщенности. Метод применяется к случаю реальных полевых данных, где, как это часто бывает в таких приложениях, данных, измеренных на скважинах, недостаточно. для правильного обучения глубоких нейронных сетей, таким образом, сеть обучается на синтетических данных. Обучение на синтетических данных дает большую свободу в разработке набора обучающих данных, поэтому важно понимать влияние распределения данных на результаты инверсии. Чтобы определить лучший способ создания набора синтетических обучающих данных, мы провести исследование четырех различных подходов к заполнению обучающей выборки. замечания о размерах данных, общности сети и влиянии физических ограничений. Используя результаты имитационной модели коллектора для заполнения наших обучающих наборов данных, мы демонстрируем преимущества ограничения обучающих выборок последовательным потоком жидкости комбинации в области динамических свойств коллектора. Благодаря этому сеть узнает физические корреляции, присутствующие в обучающем наборе, включающие эту информацию в процесс вывода, что позволяет ему делать выводы о свойствах, к которым сейсмические данные наиболее неопределенны. Дополнительно мы демонстрируем важность применения методов регуляризации, таких как добавление шума к синтетическим данным для обучения и показать возможность оценки неопределенностей в результатах инверсии с помощью обучение нескольких сетей.

bne: Именно так IMHO и выглядит конвертирование синтетики в материал обучения и характер оправданий на сей счет

B_N_E_8: Практика показывает, а логика иллюстрирует появление многоэкстремальности Но важно, что экстремумы можно разнести и оценить их число сверху В итого счет удлиняется но обозрим и можно получить матрицу экстремумов Но на что и как её употребить интересно и важно

bne: Биоинсперированные алоритмы по сути отличаются способом сдвига к экстремуму путем прогноза (за счет обобщения выбросов) Всё это невольно заставляет вспомнить книгу Катковника

B_N_E_8: Сразу две публикации В.Руденко (одна в ГЕОФИЗИКА №6 за 2020 год - видимо поздно издали) Притом на первый взгляд вполне аккуратно ссылаются (это - редкость по нынешним временам) Хотя я (как участник событий) вижу и много больше и иначе И больше западных авторов и своих публикаций Статьи можно найти по адресу https://www.elibrary.ru/cit_author_items.asp Долгие десятилетия я в СССР и в России был единственным игроком (создающим свой софт, если исключить ММ Элланского, с иными алгоритмами) Теперь должно быть веселее, поскольку есть с кем обсуждать и спорить Кстати, там же ссылки на другие цитирующие публикации (а частности про репрезентативный объём)

bne: Речь идет о суррогатном характере минимизируемого функционала Я экспериментировал с этим в 80-е годы в ЦГЭ При этом только раз написал и то мимоходом Это позволяет сократить время счёта вдали от экстремума функционала



полная версия страницы