Trading System

A sophisticated trading system with multiple strategies, risk management, machine learning capabilities, and state-of-the-art anomaly detection.

Features

• Multiple Trading Strategies
  • Bollinger Bands
  • Mean Reversion
  • MACD
  • Custom strategy support
• Advanced Anomaly Detection
  • Autoencoder-based detection
  • DBSCAN clustering
  • Isolation Forest
  • One-Class SVM
  • Local Outlier Factor
  • Ensemble detection with dynamic weighting
  • Market regime-aware signals
• Risk Management
  • Position sizing
  • Drawdown limits
  • Leverage control
  • Confidence-based filtering
• Machine Learning Integration
  • Ensemble models
  • Neural networks
  • GPU acceleration
• Real-time Monitoring
  • Web dashboard
  • Performance metrics
  • Trade logging
  • Anomaly visualization

Requirements

• Python 3.8+
• NVIDIA GPU (optional, for accelerated training)
• CUDA Toolkit 11.0+ (if using GPU)
• TA-Lib (for technical analysis)

Installation

1. Clone the repository:

git clone https://github.com/yourusername/trading-system.git
cd trading-system

2. Create and activate virtual environment:

python -m venv venv
source venv/bin/activate  # Linux/Mac
.\venv\Scripts\activate   # Windows

3. Install dependencies:

pip install -r requirements.txt

4. Install GPU dependencies (optional):

pip install -r requirements-gpu.txt

Quick Start

1. Configure your settings in config/config.yaml:

trading:
  symbols: ["AAPL", "MSFT", "GOOGL"]
  initial_capital: 100000.0
  risk_per_trade: 0.02
  max_drawdown: 0.2
  max_leverage: 2.0

anomaly_detection:
  methods: ["autoencoder", "isolation_forest", "one_class_svm", "lof", "dbscan"]
  anomaly_threshold: 0.8
  confidence_threshold: 0.7
  weights:
    autoencoder: 0.3
    isolation_forest: 0.2
    one_class_svm: 0.2
    lof: 0.2
    dbscan: 0.1

data:
  start_date: "2020-01-01"
  end_date: "2023-12-31"
  interval: "1d"

model:
  type: "ensemble"  # or "neural_net", "xgboost", "random_forest"
  use_gpu: true
  batch_size: 1024
  num_workers: 4

2. Run the training session:

python src/blueprints/training.py

3. Start the web dashboard:

python src/web/app.py

Project Structure

trading-system/
├── config/
│   └── config.yaml
├── src/
│   ├── blueprints/
│   │   └── training.py
│   ├── core/
│   │   ├── market_data/
│   │   ├── strategies/
│   │   │   ├── anomaly_detection.py
│   │   │   ├── anomaly_methods.py
│   │   │   └── ...
│   │   ├── portfolio/
│   │   └── risk/
│   ├── web/
│   │   └── app.py
│   └── utils/
├── results/
│   ├── training/
│   └── models/
├── requirements.txt
├── requirements-gpu.txt
└── setup.py

Training Process

1. Data Collection

   • Fetch historical market data
   • Calculate technical indicators
   • Prepare feature sets

2. Anomaly Detection Training

   • Train autoencoder
   • Train DBSCAN
   • Train Isolation Forest
   • Train One-Class SVM
   • Train Local Outlier Factor
   • Optimize ensemble weights

3. Strategy Testing

   • Run backtests on multiple strategies
   • Optimize strategy parameters
   • Evaluate performance metrics

4. Model Training

   • Prepare training data
   • Train selected model type
   • Validate and save results

Web Dashboard

Access the dashboard at http://localhost:8000 to:

• Monitor real-time trading
• View portfolio performance
• Analyze strategy results
• Adjust trading parameters
• Visualize anomaly detection
• Monitor market regimes

GPU Acceleration

The system supports GPU acceleration for:

• XGBoost training
• Neural network training
• Data processing
• Anomaly detection (autoencoder)

To enable GPU support:

1. Install NVIDIA drivers
2. Install CUDA Toolkit
3. Set use_gpu: true in config

Contributing

1. Fork the repository
2. Create a feature branch
3. Commit your changes
4. Push to the branch
5. Create a Pull Request

License

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

Support

For support, please open an issue in the GitHub repository.