ALGORITMA K-NEAREST NEIGHBOR (K-NN) DAN SINGLE LAYER PERCEPTRON (SLP) UNTUK KLASIFIKASI PENYAKIT ALZHEIMER

Abstract

Alzheimer's disease is a brain disorder that causes memory loss, decreased thinking skills, communication difficulties, and behavioral changes. Early detection of this disease is very important for proper treatment and planning of medical needs. However, there is currently no drug that can cure Alzheimer's. Therefore, this study aims to develop accurate early predictions for Alzheimer's disease by comparing two algorithms: K-Nearest Neighbor (KNN) and Single Layer Perceptron (SLP) using the percentage split method. The results showed that testing using the K-NN algorithm resulted in an accuracy of 96%. The precision and recall values for class 0 (nondemented) are 93% and 100%, respectively, while for class 1 (demented) are 100% and 91%. On the other hand, testing using the SLP algorithm produces an accuracy of 99%. The precision and recall values for class 0 (nondemented) are 97% and 100% respectively, while for class 1 (demented) are 100% and 98%. Based on a comparison of the values for accuracy, precision, and recall, as well as the performance of the two classification methods, it can be concluded that the implementation of the Single Layer Perceptron algorithm provides the best prediction for early detection of Alzheimer's disease. These findings provide potential use of this algorithm in facilitating early diagnosis and timely intervention for patients with Alzheimer's.