Detection of Tumor for Breast Cancer Patient Using Regression Tree Compared with Linear SVM Algorithm

Abstract

The paper aims to detect the tumor in breast cancer images using a Regression Tree and compare it with a Support Vector Machine to evaluate accuracy values.The regression tree algorithm is applied to 10 sample images from a dataset of more than 250 images. For the same input images taken accuracy, values are evaluated. The dataset taken consists of breast cancer images in this research study a Novel Regression Tree and Linear Support Vector Machine is evaluated with a total of 20 sample sizes. Based on the statistical analysis the significance value for calculating accuracy was found to be p<0.05(significant two-tailedtests). Detection of breast cancer was performed by using a Regression Tree and Support Vector Machines which achieved the accuracy of 95% and 86% respectively. By the above two results, we conclude that the performance of the Novel Regression tree is significantly better than the Linear Support Vector Machine in terms of accuracy.