Improved Method of Stock Trading under Reinforcement Learning Based on DRQN and Sentiment Indicators ARBR

TL;DR

This paper introduces an enhanced DRQN-based stock trading model incorporating sentiment indicators, which better captures investor behavior and improves trading performance in the Chinese stock market.

Contribution

It proposes a novel DRQN-ARBR model that integrates sentiment analysis and LSTM layers, addressing memory limitations and market irrationality effects in quantitative trading.

Findings

Significant performance improvement over traditional models

Effective incorporation of sentiment indicators improves decision accuracy

Model adapts well to non-efficient market environments

Abstract

With the application of artificial intelligence in the financial field, quantitative trading is considered to be profitable. Based on this, this paper proposes an improved deep recurrent DRQN-ARBR model because the existing quantitative trading model ignores the impact of irrational investor behavior on the market, making the application effect poor in an environment where the stock market in China is non-efficiency. By changing the fully connected layer in the original model to the LSTM layer and using the emotion indicator ARBR to construct a trading strategy, this model solves the problems of the traditional DQN model with limited memory for empirical data storage and the impact of observable Markov properties on performance. At the same time, this paper also improved the shortcomings of the original model with fewer stock states and chose more technical indicators as the input values of the model. The experimental results show that the DRQN-ARBR algorithm proposed in this paper can significantly improve the performance of reinforcement learning in stock trading.