Evaluation of Machine Learning Techniques for Prediction of Breast Cancer

The prediction of breast cancer prognosis has been a significant challenge in medical research due to the complex nature of cancer progression and the variability in patient-specific factors. Machine learning techniques have emerged as powerful tools for enhancing prediction accuracy in cancer diagnosis and prognosis. This study evaluates the performance of three widely used machine learning classifiers—Artificial Neural Network (ANN), Random Forest (RF), and Support Vector Machine (SVM)—in predicting breast cancer outcomes using the Wisconsin Breast Cancer Database. A robust framework was designed to preprocess the dataset, scale features, and assess the classifiers using key evaluation metrics, including accuracy, precision, recall, F1-score, and Matthews Correlation Coefficient (MCC). The results reveal that all classifiers demonstrated high predictive accuracy, with SVM achieving the highest accuracy of 97.08%, followed by RF and ANN with 96.67%. While RF recorded superior precision (95.40%), SVM outperformed others in recall (97.70%) and F1-score (96.05%), showcasing its ability to balance prediction precision and sensitivity. These findings underscore the potential of machine learning to augment traditional cancer prognosis methods, providing healthcare professionals with data-driven insights for personalized treatment strategies. This research highlights the critical role of machine learning in transforming cancer care and lays the groundwork for future studies focused on hybrid models and real-world implementation.