₿ BTC & ETH Hourly Price Prediction Model (LSTM)

A deep learning–based hourly price prediction system for Bitcoin (BTC) and Ethereum (ETH) using
LSTM neural networks, technical indicators, and multi-asset feature fusion.

The model predicts the next 1-hour closing price for both BTC & ETH simultaneously and also generates directional Buy / Sell signals with technical reasoning.

⚠️ DISCLAIMER
This project is strictly for educational & research purposes only.
It is NOT financial advice.
Crypto markets are highly volatile — do NOT use this model for live trading.

---

🧠 Core Idea

Crypto prices are influenced by:

• Momentum
• Volatility
• Trend strength
• Cross-asset correlation (BTC ↔ ETH)

This project:

• Combines BTC & ETH data
• Adds 90+ technical indicators per asset
• Uses hourly sequences (24 timesteps)
• Predicts next-hour prices using multi-output LSTM

---

🛠 Tech Stack

• Python
• TensorFlow / Keras
• LSTM (Deep Learning)
• yFinance
• ta (Technical Analysis)
• Scikit-learn
• Pandas / NumPy
• Matplotlib

---

📊 Model Architecture

• Input:
  • Last 24 hours
  • 182 features (BTC + ETH TA indicators)
• Network:
  • 3 stacked LSTM layers
  • Batch Normalization
  • Dropout regularization
• Output:
  • BTC next-hour close price
  • ETH next-hour close price
• Loss:
  • Mean Squared Error (MSE)

---

📁 Project Structure