Randomized Machine Learning Algorithms to Forecast the Evolution of Thermokarst Lakes Area in Permafrost Zones

Cover Page

Cite item

Full Text

Open Access Open Access
Restricted Access Access granted
Restricted Access Subscription or Fee Access

Abstract

Randomized machine learning focuses on problems with considerable uncertainty in data and models. Machine learning algorithms are formulated in terms of a functional entropy-linear programming problem. We adapt these algorithms to forecasting problems on an example of the evolution of thermokarst lakes area in permafrost zones. Thermokarst lakes generate methane, a greenhouse gas affecting climate change. We propose randomized machine learning procedures using dynamic regression models with random parameters and retrospective data (climatic parameters and remote sensing of the Earth’s surface). The randomized machine learning algorithm developed below estimates the probability density functions of model parameters and measurement noises. Randomized forecasting is implemented as algorithms transforming the optimal distributions into the corresponding random sequences (sampling algorithms). The randomized forecasting procedures and technologies are trained, tested, and then applied to forecast the evolution of thermokarst lakes area in Western Siberia.

About the authors

Yu. A Dubnov

Federal Research Center “Computer Science and Control,” Russian Academy of Science; National Research University Higher School of Economics

Email: yury.dubnov@phystech.edu
Moscow, Russia; Moscow, Russia

A. Yu Popkov

Federal Research Center “Computer Science and Control,” Russian Academy of Science

Email: apopkov@isa.ru
Moscow, Russia

V. Yu Polishchuk

Institute of Monitoring of Climatic and Ecological Systems, Siberian Branch, Russian Academy of Sciences

Email: liquid_metal@mail.ru
Tomsk, Russia

E. S Sokol

Yugra Research Institute of Information Technologie

Email: sokoles@uriit.ru
Khanty-Mansiysk, Russia

A. V Mel'nikov

Yugra Research Institute of Information Technologie

Email: melnikovav@uriit.ru
Khanty-Mansiysk, Russia

Yu. M Polishchuk

Yugra Research Institute of Information Technologie

Email: yupolishchuk@gmail.com
Khanty-Mansiysk, Russia

Yu. S Popkov

Federal Research Center “Computer Science and Control,” Russian Academy of Science

Author for correspondence.
Email: redacsia@ipu.rssi.ru
Moscow, Russia

References

  1. Vapnik V.N. Statistical Learning Theory. John Willey & Sons, 1998.
  2. Bishop C. Pattern Recognition and Machine Learning. N.Y. Springer, 2007.
  3. Friedman J., Hastie T., Tibshirani R. The elements of statistical learning. Volume 1, Springer series in statistics, Berlin. Springer, 2009.
  4. Popkov Yu.S., Dubnov Yu.A., Popkov A.Yu. Randomized Machine Learning: Statement, Solution, Applications // Proc. IEEE Int. Conf. on Intelligent Systems. 2016. P. 27-39.
  5. Zuidhoff F.S., Kolstrup E. Changes in palsa distribution in relation to climate change in Laivadalen, Northern Sweden, espesially 1960-1997 // Permafrost and Periglacial Processes. 2000. V. 11. P. 55-69.
  6. Kirpotin S., Polishchuk Y., Bruksina N. Abrupt changes of thermokarst lakes in Western Siberia: impacts of climatic warming on permafrost melting // Int. J. Environmental Studies. 2009. V. 66. No. 4. P. 423-431.
  7. Karlson J.M., Lyon S.W., Destouni G. Temporal behavior of lake size-distribution in a thawing permafrost landscape in Northwestern Siberia // Remote Sensing. 2014. No. 6. P. 621-636.
  8. Bryksina N.A., Polishchuk Yu.M. Analysis of changes in the number of thermokarst lakes in permafrost of Western Siberia on the basis of satellite images // Cryosphere of Earth. 2015. V. 19. No. 2. P. 114-120.
  9. Liu Q., Rowe M.D., Anderson E.J., Stow C.A., Stumpf R.P. Probabilistic forecast of microcystin toxin using satellite remote sensing, in situ observation and numerical modeling // Environment Modelling and Software. 2020. V. 128. P. 104705.
  10. Vidyasagar M. Statistical Learning Theory and Randomized Algorithms for Control // IEEE Control System Magazine. 1998. V. 1. No. 17. P. 69-88.
  11. Граничин О.Н., Поляк Б.Т. Рандомизированные алгоритмы оценивания и оптимизации при почти произвольных помехах. М.: Наука, 2002.
  12. Biondo A.E., Pluchino A., Rapisarda A., Helbing D. Are random traiding strategies more successful than tachnical ones? // PLoS ONE. 2013. V. 6. No. 7. P. e68344.
  13. Lutz W., Sandersen S., Scherbov S. The end of world population growth // Nature. 2001. V. 412. No. 6846. P. 543-545.
  14. Цирлин А.М. Методы усредненной оптимизации и их применение. М.: Физматлит, 1997.
  15. Shannon C.Communication Theory of Secrecy Systems // Bell System Technical Journal. 1949. V. 28. No. 4. P. 656-715.
  16. Jaynes E.T. Information Theory and Statistical Mechanics // Physics Review. 1957. V. 106. P. 620-630.
  17. Jaynes E.T. Papers on probability, statistics and statistical physics. Dordrecht. Kluwer Academic Publisher, 1989.
  18. Jaynes E.T. Probability Theory. The logic and science. Cambrige University Press, 2003.
  19. Попков Ю.С., Попков А.Ю., Дубнов Ю.А. Рандомизированное машинное обучение при ограниченных объемах данных. М.: УРСС, 2019.
  20. Popkov Y., Popkov A. New Method of Entropy-Robust Estimation for Ramdomized Models under Limited Data // Entropy. 2014. V. 16. P. 675-698.
  21. Иоффе А.Д., Тихомиров В.М. Теория экстремальных задач. М.: Наука, 1984.
  22. Darkhovsky B.S., Popkov Y.S., Popkov A.Y., Aliev A.S. A Method of Generating Random Vectors with a Given Probability Density Function // Autom. Remote Control. 2018. V. 79. No. 9. P. 1569-1581. https://doi.org/10.1134/S0005117918090035
  23. Айвазян С.А., Енюков И.С., Мешалкин Л.Д. Прикладная статистика: Исследование зависимостей. М.: Финансы и статистика, 1985.
  24. Электронный ресурс: https://cloud.uriit.ru/index.php/s/0DOrxL9RmGqXsV0. Статья представлена к публикации членом редколлегии А.Н. Соболевским.

Supplementary files

Supplementary Files
Action
1. JATS XML
Download

Copyright (c) 2023 The Russian Academy of Sciences