An Adaptive Neuro Fuzzy to Predict Cryptocurrency Based on the Crisp Method: Case of COVID-19

Abstract

Over the past few years, there has been a notable surge in interest towards cryptocurrency, especially in the context of the crisis, where researchers have been diligently examining the influence of the coronavirus on cryptocurrency returns.Numerous studies have utilized econometric and machine learning techniques to forecast cryptocurrency prices, but most of them have focused solely on the financial domain. This paper introduces a novel approach called ANFPC (Adaptive Neuro Fuzzy Prediction Cryptocurrency), which combines insights from both the financial and health domains to predict the price fluctuations of various cryptocurrencies such as bitcoin, ethereum, cardano, xpr, and dogecoin. The approach relies on the ANFIS Model, a fusion of fuzzy logic and artificial neural network (ANN).The experimental findings demonstrate that ANFPC provides accurate predictions, as measured by metrics like Mae and Mse, outperforming traditional ANN and LSTM methods. This approach proves to be a valuable decision support tool for data analysts in the realm of cryptocurrency prediction.