Comparative studies on breast cancer classifications with K-Fold Cross Validations using machine learning techniques

Abstract

In the past years there are several machine learning techniques have been proposed to design precise classification systems for several medical issues. This paper compares and analyses breast cancer classifications with different machine learning algorithms using k–Fold Cross Validation (KCV) technique. Decision Tree, Naïve Bayes, Neural Network and Support Vector Machine algorithm with three different kernel functions are used as classifier to classify original and prognostic Wisconsin breast cancer. The comparative analysis of the studies are focusing on the impact of k in k-fold cross validation and achieve higher accuracy. We have used the benchmarking dataset in UCI in the experiments. In theory the common choice is to select k=10 for KCV. However this comes at an increased computational cost whereby the more the folds the more models you need to train. The overall results showed important conclusion; we cannot always expect to have more accurate result by having greater value of k in k-fold cross validation.