An Enhanced Diagnostic Framework for Hepatitis C Prediction Using Machine Learning with Advanced Feature Selection Techniques

Hepatitis C is a major global health concern and one of the leading causes of chronic liver disease, liver cirrhosis, and liver cancer worldwide. In this paper, an intelligent machine-learning algorithm-based Hepatitis C diagnosis system has been proposed. A publicly available Hepatitis C dataset has been utilized to develop the prediction framework. The dataset shows a significant imbalance between positive and negative cases, which complicates the classification process, a common challenge in healthcare datasets. Appropriate strategies have been undertaken to achieve a reliable decision framework. Two feature selection techniques, Sequential Feature Selection (SFS) and Recursive Feature Elimination (RFE), were incorporated to identify the most representative set of attributes and improve the system’s diagnostic performance. Six different classifiers were employed for prediction, among which Adaboost with the SFS algorithm achieved the highest MCC score of $\mathbf{9 2. 6 \%}$, providing the most effective prediction of Hepatitis C. This framework offers a robust approach for early diagnosis, potentially benefiting patient care and optimizing healthcare resources.