Comparison of accuracy level K-Nearest Neighbor algorithm and Support Vector Machine algorithm in classification water quality status

Abstract

Water is classified into four status of water quality, which good condition, lightly polluted, medium polluted and heavyly polluted. The classification status of water quality is very important to know the proper use and handling. Accuracy in classification of the quality status is very important, so that both of the classification algorithm K-Nearest Neighbor (KNN) and Support Vector Machine (SVM) are used. The classification of status of water quality based on the parameters. This study discusses the comparison algorithm KNN and SVM in classification of water quality status, a comparison conducted to determine the value that algorithm has the highest accuracy of the determination water Quality Status Classification, testing KNN and SVM algorithm using 10-fold Cross Validation. Based on the result of the test, the highest average value of accuracy is SVM because the accuracy value is higher, it is 92.40% at linear kernel. The average value of KNN accuracy is only 71.28% at K=7.