Comparative Analysis of Machine Learning Algorithms for Breast Cancer Classification: SVM Outperforms XGBoost, CNN, RNN, and Others
Prithwish Ghosh, Debashis Chatterjee · bioRxiv (Cold Spring Harbor Laboratory) · 2024
ABSTRACT This study evaluates ten machine learning algorithms for classifying breast cancer cases as malignant or benign based on physical attributes. Algorithms tested include XGBoost, CNN, RNN, AdaBoost, Adaptive Decision Learner, fLSTM, GRU, Random Forest, SVM, and Logistic Regression. Using a robust dataset from UCI machine learning Breast Cancer, SVM emerged as the most accurate, achieving 98.2456% accuracy. While AdaBoost, Logistic Regression, Neural Networks, and Random Forest showed promise, none matched SVM’s accuracy. These findings underscore the potential of machine learning, particularly SVMs, in cancer diagnosis and treatment by analyzing physical attributes for improved diagnostics and targeted therapies.