https://github.com/mdraihansobhan/strategic-stock-trading-with-deep-reinforcement-learning-models
https://github.com/mdraihansobhan/strategic-stock-trading-with-deep-reinforcement-learning-models
a2c ddpg deep-reinforcement-learning mvo ppo sac stock-price-prediction td3
Last synced: 3 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/mdraihansobhan/strategic-stock-trading-with-deep-reinforcement-learning-models
- Owner: MdRaihanSobhan
- Created: 2024-12-14T03:34:26.000Z (6 months ago)
- Default Branch: main
- Last Pushed: 2024-12-27T14:04:58.000Z (5 months ago)
- Last Synced: 2025-01-21T03:16:43.331Z (5 months ago)
- Topics: a2c, ddpg, deep-reinforcement-learning, mvo, ppo, sac, stock-price-prediction, td3
- Language: Jupyter Notebook
- Homepage:
- Size: 136 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Strategic-Stock-Trading-with-Deep-Reinforcement-Learning-Models
## Team Members
- 1905095 : Md Raihan Sobhan
- 1905115 : Tahsin Wahid
## Problem Definition
Traditional trading models rely heavily on historical data, making them rigid and slow to respond to real-time market fluctuations. This gap often leads to missed opportunities and increased risk for traders.
## Literature Review Highlights:
- Deep Reinforcement Learning (DRL) has shown potential in enabling real-time adaptability and improving decision-making in volatile
environments.
- Studies suggest DRL can integrate real-time data and sentiment analysis, providing more robust, dynamic trading strategies.
## Datasets for Model Training
### Datasets Used:
- [Bangladesh | Dhaka Stock Exchange Dataset (DSEBD)](https://www.kaggle.com/datasets/mahmudulhaque/dsebd): Daily stock price data from the Dhaka Stock Exchange, providing insights into a growing emerging market.
- [India | NIFTY-50 Stock Market Data (2000 - 2021)](https://www.kaggle.com/datasets/rohanrao/nifty50-stock-market-data?select=BHARTIARTL.csv)
- [US | USA 514 Stocks Prices NASDAQ NYSE](https://www.kaggle.com/datasets/olegshpagin/usa-stocks-prices-ohlcv/data)
### Additional International Datasets That Can Be Used in Future Iterations:
- [SP 500 Stock Data](https://www.kaggle.com/datasets/paultimothymooney/stock-market-data): Covers daily prices for SP 500 stocks, offering a view into a mature market.
- [US Stocks and ETFs Price and Volume Data](https://www.kaggle.com/datasets/borismarjanovic/price-volume-data-for-all-us-stocks-etfs): Includes comprehensive price and volume data across all US stocks and ETFs.
- [Brazilian Stock Market](https://www.kaggle.com/datasets/andrewmvd/brazilian-stock-market Daily Updated
- Indian Stock Market Index Intraday Data(2008-2020)
## Reference Papers :
- [FinRL](https://arxiv.org/pdf/2011.09607): A Deep Reinforcement Learning Library for Automated Stock
Trading in Quantitative Finance
- [Practical Deep Reinforcement Learning Approach for Stock Trading](https://arxiv.org/pdf/1811.07522v3)
## Proposed Solution :
### Solution Approach:
- Develop a DRL-based trading model that dynamically adapts to real-time market conditions by employing an optimal reward function.
- This function not only maximizes profits but also carefully considers the level of risk taken, ensuring a balanced strategy that prioritizes sustainable growth while minimizing potential losses.
## Key Techniques:
- A2C (Advantage Actor-Critic) for efficient exploration.
- DDPG (Deep Deterministic Policy Gradient) for continuous action spaces.
- Proximal Policy Optimization (PPO) for stability in
decision-making.
- TD3 (Twin Delayed DDPG) to manage volatility and improve
accuracy.
- SAC (Soft Actor-Critic) for robustness and adaptability.
- MVO (Mean-Variance Optimization) for Bechmarking.
## Performance Metrics
- Initial and Final Portfolio Value
- Annualized Return
- Annualized Standard Deviation
- Sharpe Ratio
- Max Drawdown
- Sortino Ratio
- Volatility
- Profitability
- Aggressiveness
- Recovery Time
## How to Run This Code
- Run the jupyter notebooks , install required libraries.
- [DSE](/dse.ipynb)
- [NIFTY](/nifty.ipynb)
- [SNP_500](/snp500.ipynb)
- Update dataset paths, with your local path
### Report
- Read the [Report](/Project_Report.pdf) to get a detailed overview of the project.
## Tasks to be done
- Write Ups
- Cleandata function