diabetic-retinopathy-diagnosis
This report aims at solving the problem of Diabetic Retinopathy stage detection with machine learning techniques, particularly through CNN and SVM.
Diabetic Retinopathy presented itself as a core cause of blindness across the globe, specifically in east asian countries.There are four stages of the development of Diabetic Retinopathy. For the first three stages of non-proliferative phases, patients will be suffered from a gradual change from mild, moderate, to severe,and in the final stage, the patient would suffer from permanent loss of vision.
Looking at the results, we see that the Neural Network model including PCA has the highest accuracy of 0.74.