Forecasting Bitcoin Price by a Hybrid Structure Based on ARIMA, SVM and LSSVM Models

Bitcoin price prediction poses a considerable challenge due to its intricate, ever-changing nature, nonlinear trends and susceptibility to various influencing factors, rendering simplistic models inadequate for accurate forecasts. One of the commonly used data mining methods in the field of machine learning is the support vector machine. The purpose of this study is to assess the limitations of existing bitcoin price forecasting approaches and conventional support vector machines. Specifically, the machine’s features comprise the other seven cryptocurrency prices that exhibit a strong correlation with the bitcoin price. We suggest a combined approach of autoregressive moving average, support vector machine and the least square support vector machine model to generate fundamental predictions for the bitcoin prices. The data for the predictive features are then processed by this structure using cumulative autoregressive moving average.