Classification Model for Tumor Detection in Breast Cancer Patients Using K-Neighborhood and Decision Tree

Abstract

Breast cancer is a malignancy affecting many women worldwide and is associated with a high fatality rate. The paper's main objective is to detect the breast cancer tumor using a K-neighborhood and compare it with Decision Tree classification to evaluate accuracy using the Machine Learning technique. The k-neighborhood algorithm is applied to 10 images from a dataset of more than 300. For the same, the accuracy values are evaluated. It consists of breast cancer images in the research study of K-neighborhood machines and a decision tree with 20 sample sizes. Based on the statistical analysis, the significance value for calculating accuracy was p<0.05. Breast cancer detection is performed using a K-neighborhood, which means 87.5% and 80.5% in the Decision Tree classification. The performance of the K-neighborhood is considerably improved than the Decision Tree classification in terms of accuracy.