Multiclass Liver Disease Prediction with Adaptive Data Preprocessing and Ensemble Modeling

This paper addresses multiclass liver disease prediction, focusing on Hepatitis C and its progression through stages such as fibrosis and cirrhosis. Using the Hepatitis C dataset from the UCI repository, the authors design an adaptive data preprocessing pipeline that includes class-specific mean imputation, outlier rejection, log normalization, feature selection, feature scaling, and data balancing.

On top of this, several ensemble models are constructed by combining baseline machine learning classifiers. After rigorous hyperparameter optimization, the best model achieves training and testing accuracies of 99.87% and 99.80%, respectively, outperforming prior work on this dataset.

A user-friendly interface is also developed, enabling medical professionals to input patient information and receive automated risk assessments for liver disease stages.

Recommended citation: Abdullah Al Ahad, Bibhakar Das, Md Raihan Khan, Nitol Saha, Abu Zahid, and Mohiuddin Ahmad, "Multiclass liver disease prediction with adaptive data preprocessing and ensemble modeling," Results in Engineering, Elsevier, 2025.
Download Paper