Eye Disease Classification Using Ensemble Learning

This research presents the use of deep learning for eye diseases classification where the dataset is taken from Kaggle. This dataset comprises of 4.,217 fundus images across four categories: Cataract., Diabetic Retinopathy., Glaucoma., and Normal. Data preprocessing included resizing and augmentation techniques to enrich the training set. The proposed model uses pre-trained models such as-VGGI6., ResNet50., and InceptionV3-each enhanced with custom layers., achieving test accuracies of 87.1 %., 43.4%., and 75.3%., respectively. The ensemble model averaged predictions., yielding an accuracy of 84.0%., while the weighted ensemble model., which prioritized Inception V3.,achieved the highest accuracy of 84.7%.