A comprehensive machine learning-based swing trading system that uses technical indicators and Random Forest classification to identify profitable swing trade opportunities in financial markets.
- Machine Learning Model: Random Forest classifier with imbalanced class handling
- Technical Analysis: 100+ technical indicators including RSI, MACD, Bollinger Bands, ATR, and more
- Real-time Analysis: Live market data integration via Alpha Vantage API
- Risk Management: Automatic stop-loss and take-profit calculation based on ATR
- Backtesting: Historical performance testing with detailed metrics
- Multi-symbol Monitoring: Simultaneous analysis of multiple stocks
- Data Processing: Robust CSV/Excel data loading and preprocessing
pandas>=1.3.0
numpy>=1.20.0
scikit-learn>=1.0.0
ta>=0.8.0
joblib>=1.0.0
requests>=2.25.0
- Clone or download the system files
- Install required dependencies:
pip install pandas numpy scikit-learn ta joblib requests-
Get a free Alpha Vantage API key from alphavantage.co
-
Update the API key in the script:
ALPHA_VANTAGE_API_KEY = "YOUR_API_KEY_HERE"Create a ./historical_data/ directory and place your historical stock data files in supported formats:
- CSV files (*.csv)
- Excel files (*.xlsx)
- Parquet files (*.parquet)
Your data files must contain these columns (case-insensitive):
Dateor datetime indexOpenHighLowCloseVolume
Optional columns:
Adj Close(adjusted close price)DividendsStock Splits
Run the system:
python swing_trading_system.pyThe interactive menu provides 5 options:
- Processes historical data from
./historical_data/directory - Creates technical indicators
- Trains Random Forest classifier
- Saves model files for future use
Parameters:
swing_threshold: Minimum price movement to consider a swing (default: 0.15 = 15%)lookforward_periods: Days to look ahead for swing detection (default: 10)
- Fetches real-time data for a specific stock
- Generates buy/sell recommendation
- Provides probability scores and confidence levels
- Calculates stop-loss and take-profit levels
- Continuously monitors multiple stocks
- Alerts when high-probability opportunities are found
- Customizable check intervals and alert thresholds
- Tests strategy performance on historical data
- Provides win rate, average profit, and trade statistics
- Shows performance by exit reason (stop-loss, take-profit, max time)
- Safely shuts down the system
from swing_trading_system import SwingTradingSystem
# Initialize system
system = SwingTradingSystem(api_key="YOUR_API_KEY")
# Train model
system.train_model(
data_directory="./historical_data",
swing_threshold=0.15,
lookforward_periods=10
)
# Analyze a symbol
result = system.analyze_symbol("AAPL")
# Run backtest
backtest_results = system.run_backtest("AAPL", days_back=90)
# Monitor symbols
system.monitor_symbols(["AAPL", "MSFT", "GOOGL"], check_interval=300)SWING TRADE ANALYSIS: AAPL
Current Price: $175.23
Swing Probability: 85.2%
Confidence Level: High
🚨 SWING TRADE OPPORTUNITY DETECTED! 🚨
Recommended Stop-Loss: $168.45
Recommended Take-Profit: $184.67
BACKTEST RESULTS: AAPL
Total Trades: 12
Win Rate: 66.7%
Average Profit per Trade: 2.34%
Total Return: 31.2%
Stop-Loss Exits: 3 (Win Rate: 0.0%)
Take-Profit Exits: 5 (Win Rate: 100.0%)
Max Time Exits: 4 (Win Rate: 75.0%)
The system creates 100+ technical indicators including:
- Simple Moving Averages (5, 10, 20, 50, 100, 200 periods)
- Exponential Moving Averages (5, 10, 20, 50, 100, 200 periods)
- MACD (12/26/9)
- ADX (Average Directional Index)
- CCI (Commodity Channel Index)
- RSI (14, 21 periods)
- Williams %R
- Stochastic Oscillator
- Rate of Change (10, 20 periods)
- Bollinger Bands (20, 50 periods)
- Average True Range (14, 21 periods)
- Volatility ratios and squeeze indicators
- Volume Price Trend
- On Balance Volume
- Volume Weighted Average Price
- Volume ratios
- Doji patterns
- Hammer patterns
- Body size analysis
- Shadow analysis
# Random Forest Configuration
n_estimators = 100
max_depth = 8
min_samples_split = 20
min_samples_leaf = 10
class_weight = {0: 1, 1: 50} # Heavy weight for swing opportunities# Stop-loss and take-profit calculation
risk_reward_ratio = 0.5
minimum_stop_loss = 1% or 1 ATR
take_profit = max(swing_threshold, 2 * ATR)- More Data: Use at least 2-3 years of historical data
- Quality Data: Ensure clean, complete OHLCV data
- Parameter Tuning: Adjust swing_threshold based on market conditions
- Regular Retraining: Retrain model monthly with new data
- Risk Management: Always use stop-losses and position sizing
- The system uses class weights to handle rare swing opportunities
- Creates synthetic positive samples when needed
- Uses out-of-bag scoring for better validation
"Insufficient data" error:
- Ensure you have at least 500 data points for training
- Check that your data files have the required columns
"Missing required columns" error:
- Verify column names match expected format
- The system will attempt to standardize column names automatically
API rate limiting:
- Free Alpha Vantage API has rate limits (5 calls/minute, 500 calls/day)
- Add delays between API calls for multiple symbol monitoring
Low swing detection rate:
- Try lowering the swing_threshold parameter
- Increase lookforward_periods for longer-term swings
- Ensure sufficient historical data for training
After training, the system saves:
swing_model_enhanced.pkl- Trained Random Forest modelswing_scaler_enhanced.pkl- Feature scalerfeature_columns_enhanced.pkl- Required feature listtraining_stats.pkl- Training performance metrics
This system is for educational and research purposes only. Past performance does not guarantee future results. Always conduct your own research and consider your risk tolerance before making any trading decisions. The authors are not responsible for any financial losses incurred from using this system.
Feel free to submit issues, feature requests, or improvements to enhance the system's functionality.
This project is provided as-is for educational purposes. Please review and comply with your local financial regulations before using for live trading.