Performance Enhancement of Breast Cancer Detection using AdaBoost Ensemble based on SVM and Decision Tree

Abstract

Breast cancer is the largest cause of death in women. Breast cancer is amenable to treatment, and a favourable prognosis can be attained if the condition is detected and addressed during its initial phases. Early and accurate prediction is the foundation for effective breast cancer management and improved survival rates. Accurate diagnosis plays a crucial role in effective treatment and patient outcomes. Correct classification of malignant and benign tumours can ensure better clinical decision-making, ultimately contributing to improved patient outcomes. This research proposes a novel AdaBoost ensemble method consisting of heterogeneous support vector machine and decision tree as machine learning algorithms in base models. The proposed model can consistently deliver a high precision of 0.98, recall of 0.95, and F2 score of 0.96 scores, which indicates that the model is highly effective at correctly identifying malignant tumours while minimizing false positives. It not only surpasses traditional models but also outperforms other ensemble techniques, making it a reliable and effective tool for medical diagnostics.