Enhancing Breast Cancer Diagnosis: A Hybrid Approach Using Neural Networks and Explainable AI

Breast cancer is one of the main factors of death among women worldwide. The existing models of diagnostic models have some shortcomings regarding lack of interpretability or inadequate integration of different sources of clinical data. Such gaps are addressed by integrating machine learning (ML), neural network (NN), and Explainable Artificial Intelligence (XAI) that includes SHAP (SHapley Additive exPlanations) and LIME (Local Interpretable Model-agnostic Explanations). These tools enhance transparency by explaining what features are important to model predictions and provide local explanations for individual cases. It compares five ML models namely Logistic Regression, SVM, XGBoost, Random Forest, and k-NN, and also with MLP neural networks, giving a best accuracy of 97.8% of Logistic Regression. The performance of the model is accessed by using accuracy, precision, recall, and ROC-AUC. The framework integrates XAI to bridge the gap between predictive accuracy and clinical applicability in advancing breast cancer diagnostics.