Detection and Classification of Type 2 Diabetes using Machine Learning Techniques

Abstract

The Type 2 Diabetes has several consequences, one of which is diabetic retinopathy (DR). It typically comes from excessive glucose content in the blood arteries that supply nutrition to the retina, and it can also have many additional negative effects. In this paper, machine learning algorithms like CNN (Convolutional Neural Network), GLCM (Gray Level Co-occurrence Matrix), and Random Forest( RF) are used to predict diabetic retinopathy. In terms of performance metrices like Accuracy, Precision, Recall, and F1-score, the Convolutional Neural Network (CNN) produced the best results. Results generated by the CNN algorithm had a higher accuracy rate of 92%. The Django framework was used to develop the website, which aims to let people submit photographs and classify them into No DR, Mild DR, Moderate DR, Severe DR, and Proliferative DR categories.