Predictive Models for Liver Disease: Unraveling Patterns with Machine Learning Algorithms

Liver Disease is a widespread problem around the globe. It can give rise to many problems such as hepatitis, jaundice, fatty liver etc. Poor lifestyle and genetic factors are both contributors to it. Annually around two million people die due to liver diseases. Part of this is caused due to a lack of early diagnosis. Machine Learning approach is a feasible solution to counter this problem. In this paper we trained various machine learning models like Logistic Regression, Random Forest, Cat Boost, ADA Boost, Light GBM etc. on a dataset of Liver Disease Patients and tested the performance of the models based on various classification metrics like Accuracy, Precision, Recall, F-Score, etc. The best result was shown by XG Boost with an accuracy score of 81 percent followed by Light GBM and Cat Boost with an accuracy score of 80.0 percent and 79.9 percent respectively.