Hepatitis is an inflammatory liver disease caused by the hepatitis virus. This disease is a major health threat worldwide, including in Indonesia. The dataset has to be classified before it can be used for medical purposes. However, incomplete datasets (missing values) can impede the classification process. An approach that can be taken is by removing instances with missing values or performing imputation repairs using certain methods. In this study, there are 155 instances of hepatitis patients and two scenarios are carried out to determine the best classification accuracy. In the first scenario, naïve bayes classification is carried out by removing 75 instances with missing values. Meanwhile, in the second scenario, naïve bayes classification is carried out after performing imputation repairs with the MissForest algorithm on 75 instances of hepatitis patients out of 155 instances. Testing is conducted with data compositions of 50:50, 60:40, and 70:30. The results show that in the first scenario, the classification accuracy is 87.50%, 87.50%, and 83.33% respectively. Meanwhile, in the second scenario, the classification accuracy is 83.33%, 90.32%, and 93.62% respectively. From these results, it can be concluded that the use of the MissForest algorithm can improve performance on the accuracynaïve bayes classification.

Keywords : Classification, Missforest, Naïve Bayes