Machine learning-based chaotic dynamics prediction
Loading...
Date
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Выполнен анализ эффективности методов машинного обучения − регрессия опорных векторов и экстремальный градиентный бустинг, и некоторых известных алгоритмов оптимизации их параметров для предсказания хаотической динамики.
Разработан и реализован алгоритм оптимизации параметров на основе метода Пауэлла и алгоритма серых волков.
Реализован программный пакет для применения комбинаций изученных и разработанных методов машинного обучения и алгоритмов оптимизации их параметров.
Analysis of the efficiency of machine learning methods − support vector regression and extreme gradient boosting, in combination with well-known parameter optimization algorithms for chaotic dynamics prediction is performed. Parameter optimization algorithm based on the Powell method and gray wolf optimizer has been developed and implemented. A software package has been implemented for the use of combinations of studied and developed machine learning methods and parameter optimization algorithms.
Analysis of the efficiency of machine learning methods − support vector regression and extreme gradient boosting, in combination with well-known parameter optimization algorithms for chaotic dynamics prediction is performed. Parameter optimization algorithm based on the Powell method and gray wolf optimizer has been developed and implemented. A software package has been implemented for the use of combinations of studied and developed machine learning methods and parameter optimization algorithms.