DIABETIC RETINOPATHY CLASSIFICATION USING HYBRID SUPPORT VECTOR MACHINE-NEURAL NETWORK

Diabetic Retinopathy (DR) is a microvascular complication of diabetes which can cause perma- nent vision loss. The detection of DR is con- ducted through eye screening and it is done man- ually using qualitative scale to detect abnormali- ties on the retina. Although this approach is use- ful, the detection is not accurate; and creates an obligation for a tool that can easing the experts to classify the severity of DR. Machine learn- ing have been studied by previous researchers to propose an automatic DR classifier. However, it needs to be improvised especially in terms of accuracy