Comparison of K-Nearest Neighbour and support vector machine for choosing senior high school

Abstract

The aim of this research is to compare K-Nearest Neighbour (KNN) and Support Vector Machine (SVM) algorithm which are used for choosing senior high school recommendation. As we know that education is an important aspect in the development of a nation. The methodology that used in this research is data mining through Knowledge Discovery in Database (KDD) stages, which consist of cleaning, data integration, data selection, data transformation, data mining, pattern evaluation, and knowledge presentation. KNN and SVM are the most common algorithms used in data mining and decision support system. Either KNN or SVM in this research used for classifying the type of senior high school by input parameters, among others national examination score, student interest, and counsellor suggestion. Based on the experiment with several training and testing data, the result shows that SVM is better than KNN. SVM has an accuracy value around 97.1%, while KNN has an accuracy value around 88.5%. And also, SVM has processing time faster than KNN.