Classification of Breast Cancer Types Using Machine Learning

Abstract

Triple-negative breast cancer (TNBC) is an aggressive subtype of breast cancer (BC), lacking targeted therapies, which complicates treatment. This study applied machine learning (ML) algorithms to gene expression data to distinguish TNBC from non-TNBC. Four ML methods; Support Vector Machines (SVM), K-Nearest Neighbors (KNN), Naïve Bayes (NGB), and Decision Trees (DT) were evaluated using accuracy, recall, and specificity as performance metrics. SVM outperformed other models, achieving 90% accuracy, 87% recall, and 90% specificity. KNN followed closely with 87% accuracy, 76 % recall, and 88 % specificity, while NGB and DT exhibited lower recall rates. A 10-fold cross-validation approach confirmed the reliability of the classification results. This study demonstrates that SVM is particularly effective in distinguishing TNBC from non-TNBC using transcriptomic data, providing a promising approach for developing gene-based diagnostic tools. Future work should focus on larger datasets and refined feature selection methods to further improve performance.