Breast cancer is a leading cause of cancer deaths among women globally. Accurate diagnosis and prognosis are crucial for early treatment and survival. This study explores using machine learning classifiers and feature selection algorithms to improve breast cancer diagnosis and prognosis. Two datasets—Wisconsin Diagnostic Breast Cancer and Wisconsin Prognostic Breast Cancer—were analyzed using three classifiers (Support Vector Machine, XGBoost, and Random Forest) combined with three feature selection techniques (Recursive Feature Elimination, LASSO, and Monte Carlo Feature Selection). The results show that feature selection can enhance classifier performance while reducing the number of features used. The best-performing model was XGBoost with Recursive Feature Elimination, achieving 97% accuracy on both datasets. XGBoost with Monte Carlo Feature Selection also achieved 97% accuracy on the B dataset. These findings demonstrate the potential of optimized machine learning approaches to improve breast cancer diagnosis and prognosis.