This study explores the application of machine learning techniques in the agricultural domain, focusing on the classification of two common diseases in mango leaves: Powdery Mildew and Sooty Mould. Utilizing the MangoLeafBD dataset, the research employs a Gradient Boosting Classifier, enhanced with mean shift image segmentation and Hu moments for feature extraction. The performance of the model was rigorously evaluated through 5-fold cross-validation, yielding insights into its accuracy, precision, recall, and F1-score. The results demonstrate moderate success, with the highest accuracy and precision observed in the initial fold, indicating the model's potential for reliable disease identification. The study addresses the challenge of distinguishing between diseases with similar symptomatic appearances, offering a novel, data-driven approach for disease management in mango cultivation. This research contributes to the growing field of precision agriculture, highlighting the potential of machine learning in enhancing disease diagnosis and treatment strategies, thus supporting sustainable agricultural practices.