Machine learning techniques with ANOVA for the prediction of breast cancer

Abstract

Breast cancer is one of the most common cancer among females. In this paper, machine learning techniques are applied to a molecular taxonomy of breast cancer international consortium (METABRIC) dataset to extract prime clinical attributes. Analysis of variance (ANOVA), is used for clinical feature selection. Five different machine learning algorithms are implemented, which are support vector machine (SVM), decision tree, random forest, AdaBoost and artificial neural network (ANN). Among all the machine learning classifiers, ANN gives the highest accuracy of 87.43%. This statistical technique is helpful for the detection of breast cancer, and it will increase the survival rate of females.