Prediction of hepatitis-C virus using statistical learning models

The hepatitis-c virus (HCV) is a viral infection that targets the liver and has emerged as a significant global health concern. This study investigates the classification of HCV patients by identifying the potential factors crucial for the progression and early detection of HCV. The study includes dataset of 615 HCV patients from the UCI Machine Learning Repository for illustrative purposes and analyzed it through machine learning models such as naive Bayes (NB), random forest (RF), support vector machine (SVM), logistic regression (LR), decision trees (DT), and artificial neural network (ANN). The models were evaluated using various performance metrics, and a comparative analysis using non-parametric tests was conducted to evaluate the statistical significance of the model. The empirical findings show that the RF model achieved the highest performance, with an accuracy of 96.71% with Brier score (BS) of 0.035 and Matthews correlation coefficient (MCC) of 0.849, an accuracy of 96.45% with BS of 0.031 and MCC of 0.837 and an accuracy 97.41% with BS of 0.026 and MCC of 0.947 when evaluated using all features, using selected features, and selected features with the application of the synthetic minority oversampling technique (SMOTE). The analytical methods have improved the overall predictive accuracy for HCV infection and will aid in the early identification of the disease. As a result, patients can be treated at the earliest possible stage, thereby increasing the number of lives saved.