Deep Reinforcement Learning for Stock Trading

Overview

Integrated FinBERT for sentiment analysis in a PPO-LSTM model, achieving a 134.39% cumulative return and a 1.46 Sharpe ratio.

Highlights

  • Publication: Accepted for IEEE ICCKE 2024
  • Approach: Combined FinBERT’s sentiment analysis with PPO-LSTM, outperforming other models.
  • Impact: Demonstrated resilience in volatile markets.

View on GitHub