Enhancing Liver Disease Diagnosis: Ensemble Techniques for Predictive Modeling and Diagnostic Strategies

Computer Science

Hepatitis C is a liver illness caused by a virus that is transferred by contact with contaminated blood, most commonly through drug use and needle sharing. It can result in chronic infection, leading to serious health issues like cirrhosis and liver cancer. Symptoms may not be apparent until the disease has progressed. Prevention involves avoiding behaviors that can spread the virus, and testing is essential as treatments can cure most cases within a few months. The objective is to leverage this data to predict liver disease in patients, providing significant benefits to both medical practitioners and individuals affected by the disease. Machine learning methods are employed due to the extensive amount of data available, enabling the utilization of past data to forecast future cases. To address this, the study proposes a performance optimization strategy that considers the training data and the variables that have a significant impact on the predictive model. Logistic Regression, SVM, KNN, Random Forest, Nave Bayes, and Stacking ensemble techniques are used to train the upgraded preprocessed data. Comparative analysis is conducted on the six models and against other research models. The novel model employing stacking classifier and surpasses others, achieving remarkable testing accuracy of 96%. This showcases our approach as a practical solution for real-world liver disease detection.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.