Breast Cancer Classification using Machine Learning and Explainable AI on the Wisconsin Diagnostic Dataset

Breast cancer remains a leading cause of cancer-related mortality among women globally, underscoring the critical need for accurate and early diagnostic tools. This study presents a comparative analysis of five supervised machine learning algorithms for binary classification of breast tumors as malignant or benign using the Wisconsin Diagnostic Breast Cancer (WDBC) dataset. Logistic Regression, K-Nearest Neighbors, Decision Tree, Multi-Layer Perceptron, and Support Vector Machine models were evaluated after feature standardization and stratified train-test splitting. Logistic Regression, MLP, and SVM achieved the highest accuracy of 97.37%, demonstrating strong predictive performance. Explainable AI techniques including feature correlation analysis, logistic regression coefficient interpretation, and decision tree rule extraction were employed to enhance transparency and clinical trust. The findings demonstrate the effectiveness of machine learning and explainable AI for supporting breast cancer diagnosis.