<![CDATA[Liver disease is a serious health problem that requires early and accurate diagnosis. This study develops and evaluates a kernel-based Naive Bayes algorithm for liver disease classification, comparing it with standard Naive Bayes. A dataset from Kaggle was used, covering a wide range of medical variables. After data preprocessing, both models are trained and evaluated using standard metrics. Results show significant improvements over the kernel-based model, with accuracy reaching 99% compared to 80% for the standard model. Feature importance and learning curves analysis is carried out for deeper understanding. This study demonstrates the great potential of using kernel-based Naive Bayes in improving liver disease diagnosis, which may contribute to improved clinical outcomes and quality of patient care.]]>
Copyrights © 2024
