Machine Learning Approaches for Liver Disease Prediction: A Comparative Analysis
Srikanth Lakumarapu, R. Nithyanandhan, V. Sharmila Bhargavi, T P Anish, M. Nalini, R. Sıva Subramanıan
Abstract
ML has been helpful in the health sector and has been on the forefront in delivering progress in health domain linked with diagnoses and patients. Liver disease is one of the pressing health issues affecting different areas of the globe and treatment depends on its accurate and timely diagnosis. This research applies four algorithms namely; LR, KNN, DT, and RF that can be used to forecast liver illness. Based on the performance measure such as accuracy, precision, recall and F-measure, their effectiveness maybe be determined. Preliminary study shows good results for all methods, with Logistic Regression and Random Forest outperforming the others. Notably, both attain 74% accuracy, with Random Forest yielding the greatest F-measure score 72%. K-Nearest Neighbours and Decision Trees perform well but may need more optimization. This study shows that ML algorithms, namely Logistic Regression and Random Forest, helps in improving liver disease prediction and thereby leading to better healthcare outcomes.