Bitcoin Price Prediction Using GRU

Abstract

Cryptocurrency markets involves the buying, selling, and trading of digital currencies such as Bitcoin, Ethereum, Litecoin and others. Understanding the cryptocurrency markets is crucial for investors and traders in making right decisions. Predicting Bitcoin prices is challenging due to high volatility and complex market dynamics. If it is possible, then the investors gain more profits. This project uses machine learning algorithm GRU to analyze historical data and external factors to forecast Bitcoin prices. The dataset was collected from Yahoo Finance. The GRU models used because by comparing the performance with traditional machine learning models such as Linear Regression, ARIMA, and Random Forests. Evaluation of the models was done based on metrics such as mean absolute error (MAE), root mean squared error (RMSE), mean squared error (MSE), r2 score (R2) and accuracy to determine the effective performance of GRU in predicting Bitcoin prices.