A Comparitive Study of Clasification Liver Dysfunction with Machine Learning

This article presents the comparison and comparison of data of patients with liver dysfunction. By collecting information on liver disease and collecting data for selection in data mining. The Liver Disorders Data Set (UCI Machine Learning Repository) was used to compare the 359 patients with liver disease. The classification consisted of 7 types of liver disease and divided into 2 classes, namely, those with normal liver function and those who did not. Abnormal liver. The result was that the data was sorted using the Rules Part accuracy rate with 64.62 %. The OneR rule technique was 58.21 %. The Tree Decision Stump technique was 60.16 %. Tree REPTree has 62.67 % and Tree Random forest technique is 75.76 %. The results of this study showed that the tree random sampling technique was used to extract data from the 359 samples. The sample was extracted with 75.76 %. Because of the comparison results, Tree Random forest provides the most accurate value.