This script is a machine learning pipeline for predicting the closing price of Bitcoin using time series data. It starts by reading in a csv file containing Bitcoin data, resampling it, and splitting it into training and testing sets. Then, it fits and predicts using four different models: LinearRegression, BayesianRidge, SVR and RandomForestRegressor. The script then takes the average of these predictions, and prints the result. Next, it performs a grid search to optimize the parameters of a linear regression model, and then trains a neural network using the Keras library. Finally, it makes predictions using this neural network, prints the predictions and the best parameters and scores of the grid search.
- Python 3.x
- pandas
- sklearn
- keras
- requests
-
Clone the repository or download the script
git clone https://github.com/your-username/Bitcoin-Price-Prediction-Pipeline.git
-
Install the required packages
pip install [package]
-
Make sure you have a .csv file containing Bitcoin data in the same directory as the script. The file should have columns named "Date", "Open", "High", "Low", "Close" and "Volume".
-
Run the script by using the command:
python main.py
The script uses the provided data to train a model and make predictions. The predictions may not be accurate due to the volatility of the Bitcoin market. The script is for demonstration purpose only.