Predicting Cryptocurrency Price Using Machine Learning

Abstract

The cryptocurrency market which is based on blockchain technology permits execution of financial transactions over a peer-to-peer network. The primary feature of the cryptocurrency market is its extreme volatility and hence requires a method of predicting its price with precision. The current study aims at predicting the price of Litecoin, Ethereum, and Bitcoin, the top three cryptocurrencies. The data for the study comprises the monthly price of these three cryptocurrencies collected from popular crypto tracking websites for the period between January 2017 and December 2021. This data is analyzed using five Machine Learning models. The results suggested that accuracy outcomes are maximum using the Support Vector Machine technique with RMSE and MAPE values closest to 0.