Academic Journal of Engineering and Technology Science, 2024, 7(6); doi: 10.25236/AJETS.2024.070608.
Chen Kelin, Wen Zhanjie
College of Internet Finance and Information Engineering, Guangdong University of Finance, Guangzhou, Guangdong, China
The high volatility and nonlinear characteristics of modern financial markets pose significant challenges for stock index prediction. Traditional statistical models and machine learning methods struggle to effectively address abrupt market events and shifts in investor sentiment. To tackle this issue, this paper proposes an adaptive, high-accuracy prediction model based on Deep Reinforcement Learning (DRL). Using historical data from the S&P 500 Index and AAPL stock between 2015 and 2020, the study develops a dynamic trading strategy framework that integrates technical indicators, market sentiment analysis, and reinforcement learning algorithms such as PPO and DQN. The proposed model incorporates an adaptive mechanism to enable real-time market state identification and strategy optimization. Experimental results demonstrate the model's superior performance across key metrics, with an annualized return of 104.8%, a Sharpe ratio of 1.92, and a maximum drawdown of 31.4%, significantly outperforming traditional buy-and-hold strategies and baseline reinforcement learning models. This study highlights the model's exceptional predictive and risk management capabilities in complex market conditions, providing novel methodological support and practical insights for quantitative trading and investment strategy optimization in financial institutions.
Deep Reinforcement Learning; Stock Index Prediction; Adaptive Trading Strategy
Chen Kelin, Wen Zhanjie. Adaptive Deep Reinforcement Learning: A New Framework for Stock Index Prediction. Academic Journal of Engineering and Technology Science (2024) Vol. 7, Issue 6: 52-61. https://doi.org/10.25236/AJETS.2024.070608.
[1] Chen, J. C., Chen, C. X., Duan, L. J., & Cai, Z. (2022). DDPG based on multi-scale strokes for financial time series trading strategy. In Proceedings of the 2022 8th International Conference on Computer Technology Applications (pp. 22-27).
[2] Lee, K., Kim, S., & Choi, J. (2023). Adaptive and explainable deployment of navigation skills via hierarchical deep reinforcement learning. In 2023 IEEE International Conference on Robotics and Automation (ICRA) (pp. 1673-1679). IEEE.
[3] Hao Shuang, Li Guoliang, Feng Jianhua, Wang Ning.(2018). A Survey of Structured Data Cleaning Techniques. Journal of Tsinghua University (Science and Technology), 58(12): 1-10.
[4] Li Guoliang, Feng Jianhua. (2002). A Survey of Data Quality and Data Cleaning Research. Computer Science, 29(1): 8-14.
[5] Liao, S. (2020). A review of data cleaning research. Computer Knowledge and Technology, 16(20), 44-47.