Seeking best performance: a comparative evaluation of machine learning models in the prediction of hepatitis C

Abstract

Hepatitis C is a disease that affects millions of people worldwide. It is spread through contact with contaminated blood through injections, transfusions, or other means. It is estimated that with early detection patients have a higher rate of recovery. The objective of this study is to perform a comparative evaluation of different models focused on the prediction of hepatitis C, to determine which of the models offers better performance in accuracy, precision, and sensitivity. The models used were logistic regression (LR), random forest (RF), K-nearest neighbors (KNN), decision tree (DT), and gradient boosting (GB), aimed at hepatitis C prediction. The training of the models was carried out using a dataset composed of 615 records, which incorporate 14 attributes. The structure of the article is divided into six sections, including introduction, review of related articles, methodology, results, discussion, and conclusions. The performance of the models was evaluated through metrics such as accuracy, sensitivity, F1 count, and, mainly, precision. The results obtained place the DT model as the most efficient predictor, reaching a precision, accuracy, sensitivity, and F1-score of 95%.