Applying the Algorithm of Fuzzy K-Nearest Neighbor in Every Class to the Diabetes Mellitus Screening Model

Abstract

Heart disease is the number one killer in the world. Someone who has the potential to experience heart disease is a person with Diabetes Mellitus (DM). Diabetes that is detected early can reduce the risk of heart disease and various other complications. This study aims to analyze the performance of the model that is expected to be an alternative for screening DM by using a machine learning method, namely the Fuzzy K-Nearest Neighbour in Every Class (FKNNC) algorithm. The input in this study was clinical data from Hospital X which consisted of 7 predictor variables and 1 response variable. We used the confusion matrix and the Area Under Curve (AUC) value of the Receiver Operating Characteristic (ROC) curve to maseure the performance of the FKNNC model, with the help of the Python programming language. The results obtained indicate that the FKNNC model is classified as a “good classification” model. This can be seen from the accuracy of the FKNNC model of 86% and F1-score of 81,8%.