Open AccessEducational Data Mining in Predicting Student Final Grades
Author(s) -
William Damopolii,
Nathan Priyasadie,
Amalia Zahra
Publication year - 2021
Publication title -
international journal of advanced trends in computer science and engineering
Language(s) - English
Resource type - Journals
ISSN - 2278-3091
DOI - 10.30534/ijatcse/2021/521012021
Subject(s) - decision tree , computer science , feature selection , educational data mining , naive bayes classifier , machine learning , field (mathematics) , data mining , selection (genetic algorithm) , feature (linguistics) , k nearest neighbors algorithm , artificial intelligence , implementation , data science , support vector machine , mathematics , linguistics , philosophy , pure mathematics , programming language
Educational data mining is a field of science that extracts knowledge from educational data. One of its implementations is to predict student performance, it helps teachers to identify students that need more support. This can potentially increase learning effectiveness and elevate overall student’s grades. There are various algorithms and optimization solutions to predict student’s performance. In this paper, we use real data from one of Indonesia’s public junior high schools to compare naive bayes, decision tree, and k-nearest neighbor algorithms and implement feature selection and parameter optimization to identify which combination of algorithm and optimization can achieve the highest accuracy in predicting student grades, i.e. 7-grade classification.The results show that k-NN achieves the highest accuracy with 77.36%, where both feature selection and parameter optimization are applied