Automatic Feature Selection for Modified K-Nearest Neighbor to Predict Student's Academic Performance

Muhammad Wafi; Umar Faruq; Ahmad Afif Supianto

doi:10.1109/SIET48054.2019.8986074

Automatic Feature Selection for Modified K-Nearest Neighbor to Predict Student's Academic Performance

Closed

Muhammad Wafi, Umar Faruq, Ahmad Afif Supianto

2019 Proceedings of 2019 4th International Conference on Sustainable Information Engineering and Technology, SIET 2019 Conference paper Cited by 12

Abstract

Educational Data Mining (EDM) has been widely used in the world of education, one of them is to predict student's academic performance. There are several methods that are often used in EDM, including classification and clustering. One of the most used classification methods is K-Nearest Neighbor (K-NN), which has several modifications to improve its accuracy. To improve accuracy, we can also use the best features that can represent the whole object, therefore feature selection methods is needed to get the best results. In this paper, genetic algorithm is proposed, which have a good heuristic properties method for automatic feature selection while modified K-NN will be used as the classification method. This benefit can be used as feature selection in both K-NN and modified K-NN. The result shown that feature selection using genetic algorithm in modified K-NN is better with 82.6% accuracy than the ordinary one. © 2019 IEEE.

Affiliations

Faculty of Computer Science, Universitas Brawijaya, Malang, Indonesia

View full text (DOI) →

Research at a Glance

Research at a Glance

Topics & SDG Alignment

Collaboration

Author Profile (Selected)

References Overview

Journal & Source

Metadata & Integrity