基于强化学习的股票日内交易的研究  

Research on Stock Intraday Trading Based on Reinforcement Learning

作　　者：张晋阳 ZHANG Jinyang(College of Computer Science,Sichuan University,Chengdu 610065)

机构地区：[1]四川大学计算机学院,成都610065

出　　处：《现代计算机》2021年第14期61-64,共4页Modern Computer

摘　　要：股票日内交易由于持仓时间短,规避了一些市场风险,同时短期股票期货的价格波动主要受到投机者的影响。一些缺乏股票期货领域专业知识的用户也希望借助股票日内交易自动化系统在股市致富。为了实现这个目标,本文使用相同训练数据对Q-learning agent训练多次,然后根据不同agent生成的金融市场的交易信号来优化最终的交易决策。日内交易的实验结果与传统的买卖策略相比有更好的表现。Due to the short holding time,the stock intraday trading avoids some market risks.At the same time,the price fluctuation of short-term stock futures is mainly affected by speculators.Some users who lack professional knowledge in the field of stock futures also hope to get rich in the stock market with the help of the automatic system of day stock trading.In order to achieve this goal,we exploit a Q-learning agent trained several times with the same training data and investigate its ensemble behavior in important real-world stock markets.Experimental results in intraday trading indicate better performance than the conventional Buy-and-Hold strategy.

关 键 词：强化学习 Deep Q-learning 日内交易决策 

分 类 号：F832.51[经济管理—金融学]