Heuristic Evaluation of Computational Models in Breast Cancer Detection: An Ontological Examination of Neural and Traditional Approaches

Breast cancer remains a significant global health concern, contributing substantially to morbidity and mortality among women. The necessity for early and accurate detection is paramount, as it directly impacts treatment efficacy and patient survival outcomes. This study presents a comprehensive evaluation of five advanced machine learning (ML) and deep learning (DL) models-Random Forest, K-Nearest Neighbors (KNN), XGBoost, Convolutional Neural Network (CNN), and Recurrent Neural Network (RNN)-for the classification of breast cancer using the Wisconsin Breast Cancer dataset. The deep learning models, CNN and RNN, exhibited the highest performance metrics, both achieving an accuracy of 97.37%, coupled with Area Under the Curve (AUC) scores of 99.75% and 99.87%, respectively. XGBoost demonstrated strong per- formance across multiple metrics, achieving an accuracy of 97.36% and leading in precision with a score of 98.14%. The Random Forest and KNN models, while slightly lower in accuracy at 96.49%, maintained robust AUC scores of 99.55% and 98.06%, respectively. These findings underscore the potential of CNN and RNN models in significantly enhancing the early detection and accurate classification of breast cancer. Future research should focus on optimizing these models and validating their application across diverse clinical environments to improve diagnostic accuracy and patient outcomes.