Improving Hepatitis-C Detection Accuracy Improved C5.0 Algorithm with Chi Square Feature Selection

Hepatitis C is a serious global health problem that affects over 50 million people worldwide. One million new cases occur every year, and cirrhosis and liver cancer are associated with significant mortality rates. As the disease often progresses silently, it is essential that early detection methods are accurate. By combining advanced machine learning techniques, specifically the C5.0 decision tree algorithm enhanced with Chi-Square feature selection, this project aims to improve the detection accuracy of Hepatitis-C. Data preprocessing steps such as label encoding of categorical features and normalization of numerical data are employed to enhance the performance of the machine learning model. The Chi-Square method identifies the most relevant attributes, thereby reducing dimensionality and preventing overfitting. Using metric such as Information Gain and Entropy, the C5.0 algorithm constructs a decision tree that is optimized for classification tasks while employing aggressive pruning to improve the generalization of the model. An evaluation matrix was used to calculate metrics such as accuracy, precision, recall, and F1-score, which allowed a detailed analysis of performance. The results show that the C5.0 model outperforms standard algorithms in detecting Hepatitis-C with more accuracy.