Application of the K-Nearest Neighbor Method for Classification of Leiomyoma (Myoma)

Abstract

Information technology is very important in the process of human life. Along with the growth opens up opportunities for relatively large data growth, one of which is hospitals. Mioma is a disease that continues to increase and has a major impact on health female reproduction. Myoma is a benign tumor that grows in or around the uterus. Mioma is a medical condition experienced by women of all ages, but is often experienced by women who have entered pre-menopause, myoma is also the second benign tumor in Indonesia by age range sufferers 20-50 years old. Sufferers rarely cause specific symptoms so women are rarely aware of them the presence of myoma growth in their uterus. This research classifies patient data with a purpose to classify types of myoma disease using the K-Nearest Neighbor method. There are several The attributes used in this research are diastolic blood pressure, systolic blood pressure, hemoglobin, ever been pregnant, symptoms 1 and symptom 2. The data used for this research amounted to 288 myoma patient data which will be divided into 2, namely 70% training data and 30% testing data. Then it is divided into 3 classes, namely intramural myoma, submucosal myoma, and subserosal myoma. Results of myoma classification using the K-Nearest algorithm Neighbor at Aceh Tamiang Regional Hospital used 87 test data or patient data, indicating people with the disease Intramural myoma are more numerous with 48 data, subserosal myomas 15 data and for subserosal myomas there are 25 data with a high accuracy rate of 93%.