This study aims to investigate the feasibility of using three popular classification algorithms, namely Logistic regression, Naïve Bayes, and K-Nearest Neighbors (KNN), to predict the direction of cryptocurrency prices.
The direction of cryptocurrency prices will be classified into five categories: small increase, large increase, small decrease, large decrease, and no change. The dataset used for the study will consist of historical cryptocurrency price data, along with various technical indicators.
The performance of each algorithm will be evaluated using various metrics such as accuracy, precision, recall, and F1 score, to determine which algorithm performs best in predicting cryptocurrency price direction with five classes.
The findings of this study could potentially assist investors and traders in making more granular and informed decisions in the cryptocurrency market.