This repository contains the code for a Reinforcement Learning (RL) based asset allocation project. The project utilizes custom gym environments and advanced RL algorithms to optimize portfolio management strategies.
- Custom
PortfolioEnvenvironment for simulating portfolio optimization. - Implementation of PPO algorithm using
stable-baselines3. - GPU monitoring during training with
GPUMonitorCallback.
Clone the repository and install the dependencies:
- Prepare Data: Ensure your data is in the correct format and located at
data/processed/merged_data.csv. - Train Model: Run
train.pyto start training the RL model. - Evaluate: Use the trained model to evaluate portfolio strategies.
env.py: Contains thePortfolioEnvclass for the custom gym environment.train.py: Script to train the RL model using PPO.models/: Directory to save trained models.data/: Directory for input data.
Contributions are welcome! Please fork the repository and submit a pull request.
This project is licensed under the MIT License.
For questions, please contact [markturosy2k@gmail.com].