Estimating structure of multivariate systems with genetic algorithms for nonlinear prediction

作     者：Tomoya Suzuki Yuta Ueoka Haruki Sato 

作者机构：Department of Intelligent Systems Engineering College of Engineering Ibaraki University 4-12-1 Nakanarisawa-cho Hitachi Ibaraki 316-8511 Japan Graduate School of Engineering Doshisha University 1-3 Tatara-Miyakodani Kyotanabe-shi Kyoto 610-0321 Japan Department of Information System Design Faculty of Science and Engineering Doshisha University 1-3 Tatara-Miyakodani Kyotanabe-shi Kyoto 610-0321 Japan 

出 版 物：《Physical Review E》 (物理学评论E辑：统计、非线性和软体物理学)

年 卷 期：2009年第80卷第6期

页      面：066208-066208页

核心收录：

学科分类：07[理学] 070203[理学-原子与分子物理] 0702[理学-物理学] 

基　　金：Ministry of Education, Culture, Sports, Science and Technology Ken Millennium Corporation Grants-in-Aid for Scientific Research Funding Source: KAKEN 

主　　题：Genetic algorithms 

摘      要：Although we can often observe time-series data of many elements, these elements do not always interact with each other. This paper proposes a scheme to estimate the interdependency among observed elements only by time-series data, which is useful for selecting essential elements to optimize multivariate prediction model. Because this estimation is a sort of combinatorial optimization problems, we applied the genetic algorithm as a method to moderate this problem. Through some simulations, we confirmed performance of our method, which can identify interaction of multivariate system and can improve its prediction accuracy. Especially, our method can be applied to predict real foreign-exchange markets even if system has nonstational property and its structure changes dynamically.