Prediksi Predikat Kelulusan Mahasiswa Menggunakan Naive Bayes dan Decision Tree pada Universitas XYZ

Abstract

The success of universities in managing learning standards can be seen from the number of students who can graduate with high predicates and on time. The graduation predicate of a student is influenced by several factors. This study aims to find out the profile of students who graduated with a set predicate and what are the factors that influence it. The method used in this study is the Cross Industry Standard Process for Data Mining CRISP-DM by utilizing the Naïve Bayes algritma in looking for graduate predicate patterns and Decision Tree in looking for causative factors. In the calculations using the NBC algorithm it was found that the profiles of students who passed the predicate were less satisfactory, satisfactory, very satisfactory and with praise . In the Desicion Tree calculation, the highest gain value is obtained at the attributes of IPK4, IPS5 and IPK5. The factors that most influence graduation are the cumulative achievement index in semesters 4 and 5 and the achievement index in semester 5. The pattern of graduating with predicate can be known from the second year of the incoming student to the third year. This research needs to be developed again by increasing the number of attributes and data, and it is necessary to create a system for determining student graduation predicates from the patterns that have been produced in order to help universities to improve the quality of student graduation in each period.