📈 Crypto Price Prediction using LSTM

This project predicts Bitcoin (BTC) prices using a Long Short-Term Memory (LSTM) deep learning model. It fetches real-time Bitcoin price data from the CoinGecko API, processes it, and generates future price predictions through an interactive Streamlit web application.

---

🚀 Live Demo

🔗 Streamlit App URL 👉 https://thehoodguy4587-crypto-price-prediction-app-ytvra1.streamlit.app/

---

📌 Features

• 📊 Fetches live Bitcoin price data using CoinGecko API
• 🧠 Uses a pre-trained LSTM model for time-series forecasting
• 🔢 Data scaling using MinMaxScaler
• 📈 Visualizes Actual vs Predicted prices
• 🔮 Predicts next-day Bitcoin price
• 🌐 Deployed using Streamlit Community Cloud

---

🧠 Model Details

• Model Type: LSTM (Long Short-Term Memory)
• Sequence Length: 60 days
• Framework: TensorFlow / Keras
• Saved Format: .keras
• Scaler: Saved using pickle (scaler.pkl)

📌 The model was trained separately in Google Colab and loaded directly in the Streamlit app for inference (no retraining during deployment).

---

🗂 Project Structure

Crypto-Price-Prediction/
│
├── app.py                         # Streamlit web app
├── bitcoin_lstm_model.keras       # Trained LSTM model
├── scaler.pkl                     # Saved MinMaxScaler
├── requirements.txt               # Python dependencies
├── runtime.txt                    # Python runtime version
├── Crypto_Price_Prediction.ipynb  # Training notebook (Colab)
├── LICENSE
└── README.md

---

⚙️ Installation (Local)

1️⃣ Clone the repository

git clone https://github.com/theHoodguy4587/Crypto-Price-prediction.git
cd Crypto-Price-prediction

2️⃣ Install dependencies

pip install -r requirements.txt

3️⃣ Run the Streamlit app

streamlit run app.py

---

📦 Requirements

Key libraries used:

• streamlit
• tensorflow
• numpy
• pandas
• scikit-learn
• matplotlib
• requests

⚠️ TensorFlow version is chosen to be compatible with Python 3.12 for Streamlit Cloud deployment.

---

☁️ Deployment (Streamlit Cloud)

1. Push your code to GitHub

2. Go to 👉 https://share.streamlit.io

3. Select your repository

4. Set:

   • Main file: app.py
   • Branch: main

5. Click Deploy 🚀

---

📊 Data Source

• API: CoinGecko
• Endpoint:

https://api.coingecko.com/api/v3/coins/bitcoin/market_chart

• Currency: USD
• Time Range: Last 90 days

---

🔮 How Prediction Works

1. Fetch last 90 days of BTC prices
2. Scale prices using saved MinMaxScaler
3. Create 60-day rolling sequences
4. Predict next price using LSTM
5. Inverse transform to original price scale

---

🛠 Future Improvements

• 📅 Multi-day forecasting
• 💾 Database storage
• 📉 Technical indicators (RSI, MACD)
• 📊 Candlestick charts
• 🚀 Docker deployment

---

👨‍💻 Author

Senitha Gunathilaka 🎓 Data Science Student 🔗 GitHub: theHoodguy4587

---

📄 License

This project is licensed under the MIT License. Feel free to use, modify, and distribute.