Advanced Machine Learning Techniques for Breast Cancer Detection and Classification with Explainable AI

Breast cancer is one of the most common health issues today, and the number of cases has been very rapidly increasing over the past two decades. Machine learning is a set of technologies applied to breast cancer diagnosis, especially to make systems more accurate. Nevertheless, the uncertainty of many ML algorithms is the main challenge in healthcare, where transparency is crucial. One way to handle this matter is through the use of interpretable AI models that explain the decision-making process used by ML models to increase transparency and trust in them. The focus of this paper lies on the recent discoveries of different machine learning algorithms for breast cancer diagnosis. Finally, this study summarizes the proposed methodology for data preprocessing, training, evaluation, and interpretation. Performance metrics. Accuracy, precision, recall, precision-recall, and AUC are assessed for four ML models: Random Forest, Logistic Regression, K-Nearest Neighbors, and Support Vector Machine. The performance of the Random Forest model was the highest of all models. Further, the classifier stack and SHAP model are additionally incorporated to provide interpretability. The other research directions are the enlargement of the data sets, the exploration of cutting-edge ML methods, and the application of the models in real-time clinical decision-making to support clinicians.