Deep Q-Learning Trading Bot

Overview

This repository contains a trading bot that leverages deep Q-learning to optimize trading strategies and maximize profits. The bot is designed to interact with financial markets, continuously learning from market data to improve its decision-making capabilities over time. The goal is to achieve a higher return on investment by adapting to market conditions using reinforcement learning.

Features

• Deep Q-Learning Algorithm: The bot employs a deep Q-learning model to predict the best actions to take based on current market conditions.
• Continuous Learning: The model is trained and updated continuously to adapt to changing market dynamics.
• Profit Optimization: The bot's primary objective is to maximize profits by selecting optimal trading actions.
• Customizable Parameters: Various hyperparameters and configurations can be adjusted to tailor the bot's behaviour to specific trading scenarios.

Installation

1. Clone the repository:

   git clone https://github.com/NeuroVortex/simple-deep-q-learning-trading-bot
   cd simple-deep-q-learning-trading-bot

2. Install required packages: Make sure you have Python 3.11+ installed. Then, install the necessary dependencies using pip:

   pip install -r requirements.txt

3. Use offline Data: The bot requires offline candlestick data of specific tickers as stored in a repository.

Usage

1. Training the Model: Run the script to start training the model using historical market data:

   python train_app.py

2. Running the Bot: Once the model is trained, you can start the bot to begin trading:

   python evaluate_app.py

3. Evaluating Performance: After running the bot, you can evaluate its performance by reviewing the profit/loss logs generated in the logs/ directory.

Configuration

• Learning Rate: Adjust the learning rate in the config.py file to control how quickly the model adapts to new data.
• Exploration vs. Exploitation: Modify the epsilon-greedy strategy parameters to balance the exploration of new strategies versus the exploitation of known profitable ones.
• Batch Size: Set the batch size for training the model. Larger batches can lead to more stable learning but require more memory.
• Replay Memory: Configure the replay memory size, which stores past experiences to be replayed during training for more robust learning.

Example

Here's a basic example of how the bot might be configured and run:

from trading_bot import TradingBot

# Initialize the bot
bot = TradingBot(api_key='YOUR_API_KEY', secret_key='YOUR_SECRET_KEY')

# Train the model
bot.train(epochs=1000)

# Run the bot for live trading
bot.run(trade_duration='1d', risk_tolerance=0.05)

Contributing

Contributions are welcome! If you have ideas for new features or improvements, please feel free to submit a pull request or open an issue.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Contact

For any inquiries or questions, please contact Fatemeh Salboukh at [mkarbasioun@gmail.com].