Predicting Optimal Trading Actions Using a Genetic Algorithm and Ensemble Method

• 作者: Kazuma Kuroda
• 来源期刊: 智能信息管理(英文)
• 学科: 工学
• 关键词: Artificial INTELLIGENCE ENSEMBLE Learning Genetic ALGORITHMS Neural Networks FOREX
• 年，卷（期）: znxxglyw_2017,（6）
• 页码范围: 229-235
• 页数: 7页
• 分类号: TP39

摘要

Machine learning has been applied to the foreign exchange market for algorithmic trading. However, the selection of trading algorithms is a difficult problem. In this work, an approach that combines trading agents is designed. In the proposed approach, an artificial neural network is used to predict the optimum actions of each agent for USD/JPY currency pairs. The agents are trained using a genetic algorithm and are then combined using an ensemble method. We compare the performance of the combined agent to the average performance of many agents. Simulation results show that the total return is better when the combined agent is used.