Kombinasi Local Outlier Factor dan Winsorization dalam Pengelompokan Outlier pada Data Medis Rumah Sakit

Abstract

In medical data, it is important for doctors to stay update about the condition of their patients, but it would make a problem for the medical data. The medical data would be inconsistend due to changes in information regarding the result of examinations and leads to data that contains noises and outliers. But in medical data, the data cannot always be mistaken as noise and it cannot be deleted as it is, because the medical data might contain one or more important information. In this research, the outliers in medical data are clustered into the nearest normal cluster by using the combination between Local Outlier Factor and Winsorization, with performance level of Precision 96.8%, Recall 99.8%, Accuracy 96.9%, and F1 Score 98.3% at K = 60 which cluster as many as 47 data points.