人工市場と深層強化学習の融合による株式投資戦略学習 Learning stock trading strategy by fusion of artificial market and deep reinforcement learning

Abstract

Financial markets are known to have difficulties in predicting, such as huge elements involved, unsteady internal structure, and existence of the market impact. Even when machine learning and deep learning methods are applied, predictions must include uncertainty, and investment decision making using uncertain prediction may cause large losses and market instability. In this study, we propose to train deep reinforcement learning models in artificial market simulations for solving these problems. Artificial market simulation enables to train models in many and diverse market conditions, and also to consider market impact in training. This study provides experiments under simple market conditions, and it was confirmed that efficient strategies were learned using the proposed framework.