Bitcoin Price Prediction Using LSTM

Project Overview

This project uses a Long Short-Term Memory (LSTM) model to predict Bitcoin prices. It covers data preparation, feature engineering, model building, training, and predictions using historical price data.

Motivation

The aim is to apply machine learning techniques to financial data, developing a model that accurately predicts Bitcoin prices, potentially aiding in investment decisions.

Structure

The repository includes scripts for each stage of the machine learning pipeline and a main script to run the entire process:

• data_preparation.py: Processes raw Bitcoin price data.
• feature_engineering.py: Enhances data with technical indicators and normalizes it.
• model.py: Defines the LSTM neural network architecture.
• train.py: Trains and evaluates the model with historical data.
• predict.py: Uses the trained model for future price predictions.
• main.py: Automates the running of scripts for training and prediction.

Data

Data is downloaded from Yahoo Finance using yfinance. The raw data is stored in data/raw_data.csv.

Processed data files:

• processed_data.csv: Data with added technical indicators.
• scaled_data.csv: Normalized data for model training.

Usage

Run the entire process with main.py or individual scripts:

• Using main.py (automated process):

python main.py --train   # For training
python main.py --predict # For predictions

• Using individual scripts:

python data_preparation.py
python feature_engineering.py
python train.py
python predict.py

Optional Arguments

Run individual scripts with different ticker symbols:

python data_preparation.py --ticker ETH-USD
python feature_engineering.py --ticker ETH-USD
python train.py --ticker ETH-USD
python predict.py --ticker ETH-USD

Logging

The project uses Python's logging module to output logs to both the terminal and log files, aiding in monitoring and debugging.

Requirements

pandas~=2.1.3
numpy~=1.26.2
tensorflow~=2.15.0
matplotlib~=3.8.2
scikit-learn~=1.3.2
keras~=2.15.0
joblib~=1.3.2
yfinance~=0.2.32

Install Dependencies

pip install -r requirements.txt

Contributing

Contributions to improve the model or add new features are welcome. Please use the standard pull request process.

License

MIT