A professional-grade Bitcoin price prediction tool using ensemble machine learning models (XGBoost, Random Forest, and Gradient Boosting) with technical analysis indicators.
- Ensemble ML Models: Combines XGBoost, Random Forest, and Gradient Boosting for robust predictions
- Technical Analysis: RSI, MACD, Bollinger Bands, Moving Averages
- Real-time Data: Fetches live BTC/USDT data from Binance API
- Multiple Timeframes: 5min, 30min, 1h, 4h, 1d
- Live Auto Mode: Continuous prediction with customizable refresh intervals
- Visualization: Interactive price charts with technical indicators
- Signal Generation: STRONG BUY/BUY/SELL/STRONG SELL/HOLD signals
# Clone the repository
git clone https://github.com/codingcreatively/BTCDump.git
cd BTCDump
# Install required dependencies
active venv
pip install -r requirements.txtRun the tool:
python3 BTCDump.py- Select Timeframe - Choose prediction interval (5min to 1day)
- Train & Predict - Fetch data, train models, and get prediction
- Show Live Chart - Display price chart with technical indicators
- Refresh Data Only - Update market data without retraining
- Show Last Prediction - Display previous analysis results
- Auto Live Mode - Continuous predictions with custom refresh interval
The tool calculates the following technical indicators:
- Moving Averages: MA5, MA20, MA50
- Relative Strength Index (RSI): 14-period RSI
- MACD: 12/26 EMA with 9-period signal line
- Bollinger Bands: 20-period SMA with 2 standard deviations
- Volume Analysis: Volume SMA and ratio
The tool uses a weighted average of three models:
-
XGBoost Regressor
- n_estimators: 200
- learning_rate: 0.05
- max_depth: 6
-
Random Forest Regressor
- n_estimators: 200
- max_depth: 10
- min_samples_split: 5
-
Gradient Boosting Regressor
- n_estimators: 200
- learning_rate: 0.05
- max_depth: 5
- Current price and volume
- RSI, MACD values
- Moving averages (MA5, MA20, MA50)
- Bollinger Band positions
- Volume ratios
| Condition | Signal |
|---|---|
| Change > 1.5% AND RSI < 70 AND MACD > Signal AND Price > MA20 | STRONG BUY |
| Change > 0.5% AND RSI < 65 AND MACD > Signal | BUY |
| Change < -1.5% AND RSI > 30 AND MACD < Signal AND Price < MA20 | STRONG SELL |
| Change < -0.5% AND RSI > 35 AND MACD < Signal | SELL |
| Otherwise | HOLD |
The tool uses Binance's public API endpoint:
https://api.binance.com/api/v3/klines
No API key required for public market data.
| Option | Timeframe | Description |
|---|---|---|
| 1 | 5m | 5 Minutes |
| 2 | 30m | 30 Minutes |
| 3 | 1h | 1 Hour |
| 4 | 4h | 4 Hours |
| 5 | 1d | 1 Day |
╔════════════════════════════════════════════════════════════════════╗
║ TOOL : DUMP BTC ║
║ AUTHOR : alexxx ║
║ INSTAGRAM : arcane.__01 ║
╚════════════════════════════════════════════════════════════════════╝
BTC ANALYSIS
======================================================================
Current Price: $43,250.00
AI Prediction: $44,100.00
Change: +1.97%
RSI: 58.3
Signal: STRONG BUY
Model Error: 2.34%
Timeframe: 1h
Last Update: 2026-03-19 14:30:00
======================================================================
BTCPredictorPro
├── fetch_data() → Get BTC/USDT data from Binance
├── indicators() → Calculate technical indicators
├── features() → Create feature matrix
├── train() → Train ensemble models
├── predict() → Generate price prediction
├── signal() → Generate trading signal
├── chart() → Display price chart
├── auto_live() → Continuous predictions
└── main_menu() → User interface
- Model Error: Typically 1-3% MAPE on test data
- Training Time: ~10-30 seconds per model
- Prediction Speed: <1 second per prediction
- Data Requirements: Minimum 100 candles for training
- Cryptocurrency trading involves significant risk
- Past performance does not guarantee future results
- Do not use this tool for actual trading decisions
- The author assumes no liability for financial losses
- GitHub: @codingcreatively
- Instagram: @arcane.__01
- Binance API for real-time market data
- Scikit-learn, XGBoost, and Pandas communities
- Technical analysis principles from traditional finance