Clone the git repo, then install the requirements with pip
git clone https://github.com/shin2suka/RL-Trading.git
cd RL-Trading
pip install -r requirements.txt
Run the trading system
python backtest_main.py
Current algorithm: Deep Q learning
Urgent:
- Deploy CNN network
- Design a bette reward mechanism and the state space (NEED TO FIX profit_per_episode plot)
- Simulate several scenarios (e.g., up/down trend and mean reversion)
- Incorporate with performance engine to track stats
Non-urgent:
- Add more RL algorithms (policy gradient/PPO/A3C)
- Incorporate with dash visualization
- Build risk engine to control risk
- Build trading signals
- enable to hold multiple assets
Algorithm Trading using Q-Learning and Recurrent Reinforcement Learning
Learning to Trade via Direct Reinforcement
©2020 All Rights Reserved by Hongyi (Henry) Wu & Tianyu (Shin) Ren & Zhizu (Zale) Li