AKURASI KESAMAAN KELOMPOK DATA BERDASARKAN FCM DAN PCA-FCM PADA DATA GULA DARAH HASIL PEMINDAIAN NIRS TERHADAP DATA GULA DARAH HASIL GLUKOMETER

Abstract

Diabetes is a metabolic disorder characterized by high blood sugar levels. People with diabetes usually use a glucometer to check blood sugar levels by taking blood samples. Still, another alternative is to check blood sugar levels without using a syringe, the NIRS (Near Infrared Spectroscopy) device. In this study, data were collected using the NIRS device with blood and finger scans on three volunteers and produced the NIRS Blood Sugar Data output in the form of spectrum data. NIRS blood sugar data were grouped based on the similarity of characteristics between these objects using the Fuzzy C-Means (FCM) method. The researcher analyzed NIRS blood sugar data using PCA to reduce variables. So, the researcher only obtains significant variables in the study. Then Clustering was performed on PCA results using FCM. Based on the groups formed in each analysis using FCM and PCA-FCM, the accuracy of the similarity of data from the NIRS scanned Blood Sugar Data cluster to the Glucometer data was sought by comparing the cluster members. After grouping all the data, an accuracy rate of 50% was obtained. It shows that the Blood Sugar Data from the NIRS scan has a reasonably high similarity accuracy to the Glucometer data.