Perbandingan Algoritma Klasifikasi Naive Bayes, Nearest Neighbour, dan Decision Tree pada Studi Kasus Pengambilan Keputusan Pemilihan Pola Pakaian

Abstract

Data mining is a process of analysis of the large data set in the database so that the information obtained will be used for the next stage. One technique commonly used data mining is the technique of classification. Classification is an engineering modeling of the data that has not been classified, to be used to classify new data. Classification included into any type of supervised learning, meaning that it takes the training data to build a model of classification. There are five categories of classification that is statistically based, distance-based, based on the decision tree, neural network-based and rule-based. Each category has many options classification algorithms, some algorithms are frequently used algorithms Naive Bayes, nearest neighbor and decision tree. In this study will be a comparison of the three algorithms on case studies of electoral decision making clothing patterns. The comparison showed that the decision tree method has the highest level of accuracy than Naive Bayes algorithm and nearest neighbor, reaching 75.6%. Decision tree algorithm used is J48 with pruned algorithm that produces models of decision tree with leaves as many as 166 and 255 magnitude decision tree.