A Comparison between KNN and SVM for Breast Cancer Diagnosis Using GLCM shape and LBP Features

This paper deals with Breast cancer diagnosis from given mammogram images. Initially, the input image is being pre-processed and then features are extracted from it for the further classification. Noise and other artifacts are removed using a 2D median filter, then the features are extracted using the Local Binary Pattern (LBP) and Gray Level Co-occurrence Matrix (GLCM) feature extraction methods. Ten different features are collected from the given input image using which a Feature vector is framed. This feature vector is taken care of as a contribution to the classifiers. The classifiers used in this paper are Support Vector Machine (SVM) and K-Nearest Neighbour(KNN). To our knowledge there was no combination of features which we used were used in any of the works before. A correlation of these two classifiers are done and accuracy of 96% and 100% is acquired for SVM and KNN individually. The input data for this is taken from the CBIS-DDSM dataset.