Applying statistical learning theory to portfolio optimization problems

2014 Fiscal Year Final Research Report

• Project/Area Number: 23710174
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Social systems engineering/Safety system
• Research Institution: The University of Tokyo (2013-2014); Keio University (2011-2012)
• Principal Investigator: TAKEDA Akiko 東京大学, 情報理工学(系)研究科, 准教授 (80361799)
• Project Period (FY): 2011-04-28 – 2015-03-31
• Keywords: ポートフォリオ最適化問題 / 金融リスク尺度 / 統計的学習 / ロバスト最適化 / サポートベクターマシン

Outline of Final Research Achievements

In this study, we aimed to explore different but related research areas (mathematical optimization, financial engineering and machine learning). Before this study, we have incorporated risk measures, which are investigated in financial engineering (mainly, VaR and CVaR), into machine learning problems. On the other hand, this study brought generalization theories, which are developed in machine learning, into financial engineering based on an expectation that we might obtain a new portfolio optimization model having high prediction accuracy. Indeed, we could construct such a new portfolio optimization model which has a theoretical guarantee on high prediction accuracy and devise an efficient algorithm to solve the optimization model. Our numerical experiments supported the theoretical guarantee of the proposed model.

Free Research Field

数理最適化