STUDY TIME CLASSIFICATION OF MATHEMATICS AND INFORMATION TECHNOLOGY DEPARTMENT OF KALIMANTAN INSTITUTE OF TECHNOLOGY USING NAÏVE BAYES ALGORITHM

Abstract

Institut Teknologi Kalimantan (ITK) is one of the state universities in Indonesia which has 5 majors, one of them is the Department of Mathematics and Information Technology (JMTI). JMTI has six study programs, and only three study programs have graduates, namely Mathematics, Information Systems, and Informatics. Every year the number of new students continues to grow, but this is not proportional to the number of graduates, because some students study for more than 8 semesters. Because of this, the quality of study programs being poor. In this research, a model was built that could classify student study timeliness, using the naïve Bayes algorithm. The data used is data from JMTI student graduates from the 2013 to 2019 batch. The 2013 to 2018 batch data will be training data and validation data, while the 2019 batch data will be testing data. This research compare accuracy and F1-score naïve Bayes algorithm without correlation and with correlation. The best model obtained from training data is a model with variables that have gone through a correlation test, namely 70:30, 80:20, and 90:10. The attributes selected after the correlation test, namely, IP Tahap Bersama, GPA, Final GPA, Length of Study (Semester), dan Graduation GPA (Category), yield results for accuracy and an F1-score of 1.