Predictive Modelling for Liver Disease Diagnosis

Abstract

In this project, patient data sets are investigated to see if it is possible to predict whether a subject will acquire liver disease using simply a rigorous classification model. Since there are already procedures in place to analyze patient data and classifier data, the most crucial aspect in this situation is to anticipate the same decisive outcome with a higher level of accuracy. Recent investigations on the diagnosis of the liver revealed differences in the classification accuracy of different classifiers using varied data sets. The K-Nearest Neighbor Classifier is seen to be providing the best outcomes with India's complete feature set of liver patient data combinations. For the India liver dataset, performance is improved in comparison to the complete UCLA liver dataset and particular algorithms. In order to understand the cause. We recommend examining the liver to account for this disparity. Patients from both India and the USA To date, thorough ANOVA and MANOVA analyses have been performed on these data sets to spot any notable variations between the groupings. It has been noted that people with liver problems. The two nations have a lot of differences.