Diabetic Retinopathy’s (DR) is an eye-related infection, which is produced by an extra amount of sugar in retinal blood vessels that hinders vision. Early detection as well as prompt diagnosis reduces the severity of DR. To diagnose DR, researchers have worked on automating the classification of the illness using Machine Learning (ML) in addition to Deep Learning (DL) techniques. The main focus of this review is the research articles from 2021 to 2025 on the Support Vector Machine, K-Nearest Neighbour, as well as classification datasets. ML techniques by means of, Naive Bayes, Random Forest, Gradient Boosting classifier, Multi-class feature Extraction deep Forest are compared for DR analysis. Likewise, DL techniques such as, Convolutional Neural Network-Diabetic-Retinopathy-Estimation, CNN- Radial Basis Function, SVM- Deep Neural Network, Hybrid CNN- Singular Value Decomposition, CNN-DenseNet 121, Inception-ResNet-v2, DenseNet 121, VGG-16 and ViT –CNN. This assists the researchers in selecting the most appropriate pre-trained classification model for the relevant DR Dataset. Goal of this article is to provide researchers with access to the latest advancements in DR classification research.