Overview

Original Research

AN EFFICIENT SUPPORT VECTOR MACHINE BASED BREAST CANCER DETECTION MODEL

TSEHAY ADMASSU ASSEGIE, SUSHMA S J, BHAVYA B.G, PADMASHREE S

Vol 16, No 03 ( 2021 )   |  DOI: 10.5281/zenodo.6553289   |   Author Affiliation: Lecturer, Department of Computing Science, Faculty of Computing Technology, Aksum Institute of Technology, Aksum University, Axum, Ethiopia; Associate Professor, Department of Electronics and Communication Engineering, GSSS Institute of Engineering and Technology for Women, MYSURU, India; Assistant Professor, Computer Science and Engineering, Vidyavardhaka College of Engineering, Mysore, India; Professor, Department of Electronics and Communication Engineering, GSSS Institute of Engineering and Technology for Women, MYSURU, India   |   Licensing: CC 4.0   |   Pg no: 79-86   |   Published on: 22-03-2021

Abstract

In this research, a grid search approach is employed to develop an improved support vector machine (SVM) based breast cancer detection model. The grid search is used to find the best combinations of parameters that could maximize the accuracy of the SVM algorithm on breast cancer detection. Moreover, this study explores the effect of parameter tuning on the performance of SVM algorithm on breast cancer detection. The findings of this research shows that parameter tuning has a significant effect on the performance of the SVM algorithm on breast cancer detection. The effect of parameter tuning on the performance of SVM algorithm is experimentally tested on the Wisconsin breast cancer data collected from kaggle data repository. We have compared the performance of the SVM algorithm with the tuned hyper-parameters in the training and with default parameters. The result analysis on the performance of the SVM algorithm for breast cancer detection on the test dataset shows that the accuracy of the proposed enhanced model is 94.07% and the performance of the algorithm with the default hyper-parameters is 89.4%.

Keywords

parameter tuning, breast cancer, breast cancer detection, model optimization, parameter tuning.