This research presents a comprehensive approach for predicting stages of liver cirrhosis by integrating many learning methodologies. Utilising decision trees, random forests, and additive tree methods, we examine the feasibility of integrating several models to improve prediction accuracy. Our results indicate significant advancements in methodologies, with decision trees attaining 59% accuracy, random forests getting an impressive 90% accuracy, and additive trees reaching a maximum of 94% accuracy. We emphasise the effectiveness of collaborative learning in classifying cirrhosis based on clinical characteristics through systematic testing and comprehensive evaluation. This technique enhances diagnosis accuracy and promotes a more profound understanding of illness progression and treatment planning. The objective of this research study is to utilise machine learning skills to provide medical practitioners with effective tools for early identification and intervention, ultimately improving patient outcomes and optimising the allocation of medical resources. This study enhances the current understanding of predicted liver disease testing and paves the way for future developments in precision and personalised treatment.