This repository has the RL framework for performing futures trading in stock market. The trading environment currently supports trading in near month contracts. Discrete action space is supported for Q learning, SARSA and DQN. Continous action space is supported for PPO. Continuous observation space is supported for DQN and PPO. Discrete obeservation space is supported for Q learning and SARSA.
- deep_rl_trading: This folder consists of the PPO and DQN based RL models.
- rl_trading: This folder consists of the Q learning and SARSA based RL models.
- Price data: https://www.kaggle.com/datasets/nishanthsalian/indian-stock-index-1minute-data-2008-2020
- News data: https://economictimes.indiatimes.com/archive.cms?from=mdr
- Total Profit
- Return (%)
- Maximum Drawdown
- Volatility
- Sharpe Ratio
- Sortino Ratio
- stable-baselines3==1.7.0
- TA-Lib==0.4.25
- rich==13.3.1
- torch==1.13.0
- Futures trading: https://zerodha.com/varsity/module/futures-trading/
- Technical Indicator: https://www.fidelity.com/learning-center/trading-investing/technical-analysis/technical-indicator-guide/overview
- FinBERT: https://arxiv.org/abs/1908.10063
- PPO: https://spinningup.openai.com/en/latest/algorithms/ppo.html
- DQN: https://arxiv.org/abs/1312.5602
- Q Learning: https://blog.floydhub.com/an-introduction-to-q-learning-reinforcement-learning/
- SARSA: https://builtin.com/machine-learning/sarsa