An Analysis of theComparative Method of Classificationin Determining Characteristics of Non-Active Students

Abstract

Classification is a data mining technique that aims to know the data model being used. By using a data mining classification technique, we can classify existing information into their classes. Classification methods can also be applied in education, for example to group active and inactive students into higher education programs, and classify them based on their characteristics. This paper presents a comparison of several classification methods, which are: Naïve Bayes, k-NN and C 4.5. This paper uses the data from the inactive students of AMIK DCC campus C on 2013-2016 periods as the criteria to evaluate the group performance. The inactive students are divided into three groups: first-year, first-two-years, and three years inactive students. The results of this study indicate that the Bayes naive method provides higher accuracy than k-NN and C 4.5. The accuracy classification is Naïve Bayes 79.22%, while k-NN and C 4.5 are 77.21% and 74.94%, respectively.

Key word: Decision Tree; C4.5; classification; Naïve Bayes; AMIK DCC Campus C