Klasifikasi dengan Analisis Komponen Utama Kernel

Abstract

Principal component analysis (PCA) is a special case of the kernel PCA with linear kernel function. The aim of this study is to resolve the data problem that is not linearly separated and to classify an object into a group by using kernel PCA to obtain the smallest classification error. Group classification using kernel PCA is performed by the linear and Gaussian kernel function. The result of the study shows for wine recognition data with the linear and Gaussian (parameter 2.5) kernel function produces 30.889% and 17.416% classification error, respectively