Classification of Diabetics in Jakarta with Canonical Discriminant Analysis
Widyanti Rahayu, Vera Maya Santi, Bety Suryani Putri
Universitas Negeri Jakarta
Abstract
Diabetes is a chronic disease caused by unhealthy habits and lifestyle. Diabetes is characterized by increasing blood sugar levels in the body. Someone who has a blood sugar level slightly above normal can-t be said as a person with diabetes, such conditions are called prediabetes. Classifying diabetes case accurately is an important thing because the occurence of complications and also can be obtained significant variables that affect changes in a persons blood sugar levels. One of the statistics methods which can analyze classification of disbetes case is using linear discriminant analysis. However, when no assumptions can be made about the distribution within each group, or when the distribution is assumed not to be multivariate normal, canonical discriminant analysis can be used to classification of diabetes. In this study we will compare the classification of diabetics using linear discriminant analysis and canonical discriminant analysis. From the results, it was found that the classification accuracy with canonical discriminant analysis was higher than linear discriminant analysis
Keywords: discriminant analysis, diabetes, classification accuracy
Topic: Mathematics