Machine Learning-Based Early Detection of Liver Diseases Using Clinical and Biochemical Data

Abstract

Liver disease is a continuously increasing health concern worldwide. Liver diseases often develop without the occurrence of noticeable symptoms until the late stages. Hence, the early identification of the condition is essential for the proper therapy and management of the condition. In this study, a machine learning-based approach is proposed for the early identification of liver disease on the Indian Liver Patient Dataset. A structured preprocessing approach is implemented for the improvement of the data's quality. In this concern, various supervised ML models, including Logistic Regression (LR), Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), k-Nearest Neighbors (KNN), and Naïve Bayes (NB), are implemented for the diagnosis of the condition. To improve the performance of the presented approach, the RF-based model is tuned employing a grid search strategy. The findings have shown that the RF model has the greatest performance with an accuracy of 95.96%, along with good precision of 95.82%, recall of 94.67%, and an F1-score of 95.24%. It has been concluded from the comparative analysis of the results that ensemble learning is able to learn complex relationships between clinical data with good performance. The results of this research have shown the possibilities for using ML methods for early and accurate identification of liver disease.