<p>Glaucoma is a progressive optic neuropathy and a leading cause of irreversible blindness worldwide, with its prevalence projected to exceed 116 million cases by 2040, particularly affecting aging populations and high-risk ethnic groups. Despite advancements in diagnostic techniques, early detection remains a challenge due to the disease’s asymptomatic progression in its initial stages. Machine Learning (ML) offers a promising approach for automated, cost-effective glaucoma screening and diagnosis. In this study, we evaluate the efficacy of seven ML classifiers K-Nearest Neighbors (KNN), Random Forests, Support Vector Machine (SVM), Gradient Boosting, Decision Tree, Linear SVM, and Logistic Regression using a Glaucoma Diagnostic dataset consisting of 200 fundus images, later augmented to 2,000 images to enhance model generalizability. Performance metrics, including accuracy, precision, recall, F1-score, AUC-ROC, and confusion matrix analysis, were employed to assess classifier effectiveness. Logistic Regression demonstrated the highest diagnostic accuracy (92.8%) and AUC-ROC (0.98), outperforming other classifiers in distinguishing glaucomatous from normal cases. These results suggest that ML-driven diagnostic systems can serve as a valuable adjunct in clinical ophthalmology, facilitating early detection, risk stratification, and timely intervention to mitigate vision loss. The findings underscore the potential of ML models in improving the accuracy of glaucoma diagnosis and highlight the need for further research with more diverse datasets to ensure model generalizability and clinical applicability.</p>