UTILIZATION OF K-NEAREST NEIGHBOR ALGORITHM TO ANALYZE AND CLASSIFY HEART DISORDERS BASED ON ELECTROCARDIOGRAM RECORDING DATA

This study develops a system to classify heart conditions based on electrocardiogram (ECG) medical records using the K-Nearest Neighbor (KNN) method. This system aims to assist medical personnel, especially doctors, in analyzing ECG results more efficiently, considering the limited number of doctors and practice schedules, with the KNN method, the system can classify heart conditions based on the proximity of the patient's ECG data to other ECG data whose conditions are already known. The results of this study have an accuracy of 80%, a value of 0.88 on the Success Rate and 0.54 on Kappa. This study provides a significant contribution in the use of technology to improve the efficiency of heart examinations. This KNN-based system can be used as a tool in the diagnostic process, considering the limited medical resources. In the future, the development of this system can be done by increasing the amount of data, more complete features, or trying other more complex classification methods to improve accuracy and Kappa. Keyword: Heart Disorders, Classification, K-Nearest Neighbor, Success Rate and Kappa Statistic