Predicting Bitcoin Price with the LSTM Model

Abstract

Forecasting techniques and models play a pivotal role in guiding individuals and organizations towards informed decision-making and prudent investments. The accuracy of the forecast enables successful decisions and allows investors to maximize utility. The developments in finance and finance-related technologies around the world, along with innovative financial instruments, have piqued the interest of investors. Undoubtedly, the most prominent among these advancements is Bitcoin, a product of blockchain technology. In this study, future predictions will be generated using the LSTM model, relying on the historical data of Bitcoin and crucial market predictors. Specifically, three distinct datasets will be employed, each drawing one indicator from four different indicator types. With these datasets in hand, the study aims to predict the next ten data points pertaining to Bitcoin's performance, employing the relevant methodology. This forecast serves as a guiding model for investors and as a risk reduction study, given that it encompasses 3 different management scenarios for Bitcoin.