Comparison of Machine Learning Algorithms in the Role of Hepatitis Patient Disease Classification

Abstract

Hepatitis is one of the diseases with the highest patient percentage. about a third of the world population are afflicted with hepatitis. In several cases, patients show symptoms while in the other cases, patients show no symptoms. hepatitis is commonly caused by hepatitis A, B, C, D or E virus and yellow fever virus (YFV). hepatitis can be detected through blood test. From the blood sample, we could extract information like Alanine Transferase (ALT), bilirubin, creatine, Alkaline Phosphatase (ALP), Aspartate Aminotransferase (AST) and Gamma Glutamyl Transferase (GGT) levels, the levels of these compound will be able to determine whether the patient is afflicted or not. To raise the information processing effectiveness, machine learning can be applied to help processing the information. Several algorithms like support vector machine (SVM), decision tree, K-Nearest Neighbor (KNN), Random Forest and X-Gradient Boost (XGBoost) can be used to process hepatitis data. This research is aimed to determine which algorithm has the highest accuracy in diagnosing hepatitis.