A Systematic Method for Breast Cancer Classification using RFE Feature Selection
Ravi Kumar Sachdeva, Priyanka Bathla, Pooja Rani, Vinay Kukreja, Rakesh Ahuja
Abstract
Breast cancer is among leading reasons for the deaths of women globally. Machine learning techniques can help to classify breast cancer based on some features. In order to find a systematic method for breast cancer classification, authors have compared the performance of four different classifiers: Support Vector Machine (SVM), K-Nearest Neighbor (KNN), Logistic Regression (LR), and Random Forest (RF) on Wisconsin Breast Cancer Original (WBCO) dataset. The classifiers were used alone as well as along with techniques of feature selection. The performance with regard to accuracy, specificity, sensitivity, precision, and F -Measure, was compared for both types of experiments: classification with feature selection and without feature selection. The Recursive Feature Selection (RFE) technique was applied to select promising features out of available features. There was a significant increase in the performance of classifiers after using the RFE technique. KNN with feature selection provided the highest accuracy (98.31 %) among all other classifiers.