Nonlinear time-series analysis focusing on the background dynamics and its application to financial engineering

2015 Fiscal Year Final Research Report

• Project/Area Number: 25330280
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Soft computing
• Research Institution: Ibaraki University
• Principal Investigator: Suzuki Tomoya 茨城大学, 工学部, 准教授 (70408649)
• Project Period (FY): 2013-04-01 – 2016-03-31
• Keywords: 非線形時系列予測 / 集団型機械学習 / ポートフォリオ理論 / 金融工学 / テクニカル分析 / 複雑系 / 金融情報学

Outline of Final Research Achievements

This project had aimed to improve nonlinear time-series analysis and prediction by learning the background dynamics hidden in the observed time-series data as precisely as possible. For this purpose, the time-series data was carefully observed so as not to destroy the background dynamics. However, because it is impossible to completely eliminate prediction errors, we considered them as prediction risks, and proposed how to estimate them beforehand and reduce them effectively. Since this approach based on both the prediction and its risk management can be applied to financial engineering such as the portfolio theory, this project proposed a new framework named "nonlinear financial engineering" as an application of nonlinear dynamical theory.

Free Research Field

ソフトコンピューティング