Perbandingan Algoritma Stochastic Gradient Descent dan Naïve Bayes Pada Klasifikasi Diabetic Retinopathy

Abstract

The purpose of this research is to compare the performance of the Stochastic Gradient Descent and Naïve Bayes algorithms in classifying Diabetic Retinopathy. Diabetic retinopathy is a complication of diabetes that causes damage to the retina of the eye. These disturbances can be detected by early detection through data extracted from eye images. This research uses source data from the UCI Machine Learning Repository, namely Diabetic Retinopathy Debrecen, totaling 1,151 data records with 19 attributes consisting of 18 attributes and 1 target attribute. The validation test uses the Cross Validation method with a total of 10 k. From the comparison of the two proposed methods, the Stochastic Gradient Descent algorithm produces an average test accuracy of 70.16%, while Naïve Bayes produces an average accuracy of 56.74%. From the comparison of the two algorithms, the Stochastic Gradient Descent algorithm is known to be superior in classifying the Diabetic Retinopathy Debrecen Dataset.