Evaluation of Machine Learning Algorithms in Comparison for Early Breast Cancer Identification

Abstract

Breast cancer is a significant worldwide public health concern, and increasing patient survival rates requires early identification and precise prognosis. This research investigated a dataset on breast cancer to assess and compare the efficacy of three machine learning algorithms in the prediction of breast cancer: Support Vector Machine (SVM), Naive Bayes (NB), and Logistic Regression (LR). The findings demonstrated that the LR method provided good interpretability and computational efficiency, the NB algorithm excelled in computational efficiency, and the SVM algorithm performed best when processing high-dimensional data. Based on the findings, the paper recommend prioritizing the SVM algorithm for scenarios requiring high accuracy in breast cancer detection. Future research directions include enhancing model robustness through ensemble learning and optimizing feature selection. This study provides significant insights for early breast cancer detection, offering important clinical application value.