This study presents a comprehensive evaluation of various supervised machine learning models for the automated detection and classification of retinal diseases using the Optical Coherence Tomography Image Database (OCTID). Retinal diseases, such as Age-related Macular Degeneration (AMD), Diabetic Macular Edema (DME), and Macular Hole (MH), are leading causes of irreversible vision loss, and early, accurate diagnosis is crucial for effective treatment and prognosis. Optical Coherence Tomography (OCT) has revolutionized ophthalmic diagnostics by providing high-resolution cross-sectional images of the retina. The advent of large, publicly available datasets like OCTID offers unprecedented opportunities for developing and benchmarking automated diagnostic systems. This research systematically investigates the performance of both traditional machine learning classifiers (e.g., Support Vector Machines, Random Forests) with handcrafted features and advanced deep learning architectures (e.g., Convolutional Neural Networks) on the OCTID dataset. Through rigorous experimental protocols, including standardized preprocessing and evaluation metrics, the study compares the diagnostic accuracy, precision, recall, and F1-score of these models across different retinal pathologies. Findings indicate that deep learning models generally outperform traditional approaches, demonstrating superior capability in extracting complex, discriminative features directly from raw OCT images. This comprehensive analysis provides valuable insights into the current state-of-the-art in automated retinal disease detection using supervised learning and identifies critical future directions for enhancing diagnostic precision and clinical utility.