/CoinSavvy

CoinSavvy Crypto Price Prediction Model utilizes a linear regression model to predict Bitcoin's closing prices based on historical data.

Primary LanguageJupyter NotebookMIT LicenseMIT

CoinSavvy Crypto Price Prediction Model

Overview

Welcome to the CoinSavvy Crypto Price Prediction Model! This Python script utilizes a linear regression model to predict Bitcoin's closing prices based on historical data. The model considers the open and volume features to make predictions.

How to Use

  • Clone the repository.

    git clone https://github.com/FakePickle/CoinSavvy.git
  • Install required dependencies

    pip install -r requirements.txt
  • Run the model

    cd final_model
    python final_model.py

Results:

The script will generate a CSV file (final_predicted_values.csv) containing the predicted and actual closing values. Additionally, it will display the Mean Squared Error on both the training and test sets.
A plot comparing the actual and predicted closing values will be displayed.

Parameters

  • ticker: The cryptocurrency symbol (default is "BTC-USD").
  • start_date: Start date for historical data (default is "2018-02-01").
  • end_date: End date for historical data (default is "2022-01-01").

File Structure

├── builder_models
|    ├── dataset_builder.py
|    ├── model_tester_btcusdt.ipynb
|    ├── model_tester.ipynb
|    ├── open_prediction_model.ipynb
|    ├── open_prediction_model.py
|    ├── ridge_test_model.ipynb
├── datasets
|    ├── final_predicted_values.csv
|    ├── open_predicted_values.csv
|    ├── testing_dataset.csv
|    ├── training_dataset.csv
├── final_model
|    ├── final_predicted_values.py
├── requirements.txt
├── LICENSE
├── README.md
  • final_model/final_predicted_values.py: The main script for the Bitcoin price prediction model.
  • datasets/: Directory containing the generated CSV file with predicted values.

Notes

  • The script uses historical data from Yahoo Finance (yfinance) for training and testing the model.
  • The linear regression model considers open and volume features for predictions.
  • Adjust the start_date and end_date variables in the script to fetch data for different time periods.

License

This project is licensed under the MIT License.

Author

Harsh Mistry
Vikranth Udandarao
Noel Tiju
Mehul Agrawal

Feel free to explore, experiment, and contribute to enhancing the model.

Disclaimer: Predictive models may not guarantee accurate results. Use with caution and consider additional factors for making financial decisions.

Thank you for exploring the Bitcoin Price Prediction Model! 🚀📈
Happy predicting!