PERBANDINGAN ALGORITMA MODIFIED K-NEAREST NEIGHBOR DAN NAIVE BAYES UNTUK KLASIFIKASI DATA REKAM MEDIS BERDASARKAN ICD-10 (STUDI KASUS : RSUD RAJA AHMAD TABIB KOTA TANJUNGPINANG)

Abstract

Medical record data that increases every day can experience data accumulation so that it affects the performance of medical personnel and it is difficult to know the pattern of disease tendencies suffered in community groups. Therefore, a system was created to classify medical record data based on the code ICD-10. This research aims to obtain the results of the classification of medical record data and obtain the results of a comparison between the algorithms Modified K-Nearest Neighbor and Naive Bayes based on the level of accuracy. The data used is medical record data obtained from the Raja Ahmad Tabib Hospital in Tanjungpinang City in 2019 which has variables of district, age, gender, month of patient visit and code ICD-10. The highest accuracy results from the 3 categories ICD-10 using the Modified K-Nearest Neighbor method of 70% while the Naive Bayes method of 72% obtained 64 classified categories from 89 test data and 365 datasets. So in this research it shows that the Naive Bayes method is better than the Modified K-Nearest Neighbor based on the level of accuracy