Artificial intelligence framework for enhanced liver disease classification using multi-layer perceptron and ensemble hard-voting classifier

Early diagnosis of liver disease is crucial for improving patient outcomes and reducing disease-related complications. The potential of machine learning techniques in assisting clinical decision making by automating disease prediction has been shown to be very promising. This study presents an Ensemble Voting Classifier (EVC) which combines the Random Forest (RF) and Multi-Layer Perceptron (MLP) models to classify liver disease using the Indian Liver Patient Dataset (ILPD). The dataset consists of 583 patient records of biochemical attributes from 441 male and 142 female patients. The data was preprocessed using data cleaning, normalization, handling outliers, and feature extraction. The proposed EVC model was compared with Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Logistic Regression (LR), Random Forest (RF) and MLP classifiers. The experimental results show that the proposed RF+MLP ensemble model has an accuracy of 98.33%, precision of 98.65%, recall of 98.33%, and F1-score of 98.33% which is better than the individual machine learning models. The results show that the hard voting combination of Random Forest and Multi-Layer Perceptron is more effective in classification for liver disease prediction based on structured clinical data.