Students' Timely Graduation Attributes Prediction Using Feature Selection Techniques, Case Study: Informatics Engineering Bachelor Study Program

Abstract

Timely graduation is a problem that is often experienced by study programs at higher education institutions, where several factors can be the cause. This study applies data mining feature selection techniques to analyze attributes from student academic data which are likely affecting students' on-time graduation. The feature selection techniques used are Correlation-based Feature Selection, Information Gain Based Feature Selection, and Learner Based Feature Selection. The accuracy of each feature selection method is measured using the Naïve Bayes classification algorithm. The results of the classification test using Naïve Bayes with the application of feature selection using Correlation-based Feature Selection and Information Gain Based Feature Selection get almost the same level of accuracy as the classification test using Naïve Bayes without the application of feature selection, but the application of feature selection using Learner Based Feature Selection in the Naive Bayes algorithm, when reducing the number of features there is a possibility of increasing accuracy by eliminating features that have little relevance, namely 70.06% from 66.53%.