AT vol.2 - AI-Powered Algorithmic Trading System

A production-ready, ML-enhanced algorithmic trading system for US stock markets that autonomously generates trading signals, executes trades, and manages risk using a hybrid approach combining technical analysis with machine learning.

🎯 Overview

This system combines:

• Technical Analysis: SMA, MACD, RSI, ATR indicators with multi-signal confirmation
• Machine Learning: LightGBM model that enhances signal confidence based on historical trade outcomes
• Risk Management: Multi-layered safety controls including circuit breakers, position sizing, and drawdown protection
• Web Dashboard: Django-based monitoring interface for real-time position tracking and performance analytics

🏗️ Project Structure

AT vol.2/
├── BackEnd/              # Trading engine and ML system
│   ├── main.py          # Main trading orchestrator
│   ├── modules/         # Core trading modules (20+ components)
│   ├── ml/             # Machine learning subsystem
│   ├── data_persistence/ # Database layer (SQLite)
│   └── tests/          # Test suite
├── FrontEnd/            # Django web dashboard
│   ├── trading_app/    # Main Django app
│   └── trading_web/    # Django project settings
└── requirements.txt    # Python dependencies

🚀 Quick Start

1. Setup Environment

# Create virtual environment
python -m venv venv

# Activate (Windows)
.\venv\Scripts\Activate.ps1

# Activate (Linux/Mac)
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

2. Configuration

1. Copy .env.example to .env and add your API credentials:

   • KIS API keys (for live trading)
   • Discord bot token and channel ID (for notifications)

2. Configure BackEnd/usa_stock_trading_config.yaml:

   • Set trading mode: paper (recommended for testing) or live
   • Configure symbols to trade
   • Adjust risk parameters and ML settings

3. Start Trading System

cd BackEnd
python main.py

4. Access Web Dashboard

cd FrontEnd
python manage.py runserver

The dashboard will be available locally. See FrontEnd/README.md for details.

✨ Key Features

Trading Engine

• Hybrid Signal Generation: Technical analysis + ML confidence boosting
• Dual Modes: Paper trading (simulation) and live trading (KIS API)
• Automated Execution: State machine-based order management
• Position Persistence: Automatic save/restore on restart

Machine Learning

• 22-Feature Model: LightGBM-based confidence booster
• Continuous Learning: Automated data collection and periodic retraining
• A/B Testing: Compare ML-enhanced vs. standard performance

Risk Management

• Circuit Breakers: Automatic halt on losses or daily limits
• Position Sizing: ATR-based dynamic sizing
• Drawdown Protection: Multi-layered risk controls

Web Dashboard

• Real-time Monitoring: Positions, P&L, trade history
• Performance Analytics: Win rate, statistics, charts
• Risk Dashboard: Circuit breaker events, exposure tracking

📚 Documentation

• BackEnd/README.md - Backend implementation details
• FrontEnd/README.md - Frontend setup and usage
• BackEnd/TRADING_SYSTEM_PLAN.md - Detailed architecture

🛠️ Technologies

• Backend: Python 3.11+, LightGBM, SQLite, KIS API, yfinance
• Frontend: Django, Django REST Framework
• ML: scikit-learn, pandas, numpy
• Testing: pytest

⚠️ Disclaimer

This software is for educational and research purposes. Trading involves substantial risk of loss. Always test thoroughly in paper trading mode before considering live trading. The authors are not responsible for any financial losses.

📄 License

[Add your license here]