Classification of Osteoarthritis Disease Severity Using Adaboost Support Vector Machines

Abstract

Osteoarthritis (OA) is a condition when the joint is painful due to mild inflammation that arises due to friction of the ends of the joint bone. OA is the most chronic disease and joint disability in elderly people. One way to prevent this disease is to do early detection using machine learning for classification. In this study, it was used Adaptive Boosting (AdaBoost) and Support Vector Machines (SVM) together as classifiers. The purpose of this study was to see whether AdaBoost SVM could produce good accuracy with SVM as comparison. Tests were conducted using 10% until 90% data training. Polynomial and RBF kernel were used with number of AdaBoost cycle. The highest accuracy value of SVM was 75% in 90% training data, while the highest accuracy value of AdaBoost SVM was 85,714% in 80% training data. Therefore, it could be that AdaBoost can improve the performance of SVM in classification of OA disease severity.