Performance Comparison of Ensemble-based k-Nearest Neighbor and CART Classifiers for the Classification of Adaptive e-learning User Knowledge Levels

DOI

10.2991/assehr.k.211020.037How to use a DOI?

Keywords

Ensemble-Based K-Nearest, CART, e-learning

Abstract

A person’s learning style that refers to the preferred way of learning is the basis for the development of the Adaptive Educational Intelligent Hypermedia System (AEIHS) or adaptive e-learning. By knowing specific learning styles, the system can provide recommendations and offer instructions to someone how to optimize the learning process. The right learning style is needed, because it can support the achievement of a person’s level of knowledge in learning. The problem in determining this level of knowledge is related to the performance of data mining methods as measured by algorithm performance. High performance is indicated by the optimal results of the algorithm used. On the other hand, the K-Nearest Neighbor (KNN) algorithm and Classification and Regression Trees (CART) have been shown to have good performance in various fields. Therefore, this study aims to compare the performance of data mining methods, namely K-Nearest Neighbor (KNN) and Classification and Regression Trees (CART). There are six scenarios carried out for comparison, including comparing the performance of the original algorithm and ensemble methods, namely bagging and boosting. The results of this study are the best performance results for the classification of the level of user knowledge of adaptive e-learning from several scenarios performed. CART boosting algorithm shows the best performance with accuracy of 94.0%, precision 94.2%, and recall of 94.0%. The best scenario of the algorithm used is expected to be a guide for developing AEIHS-based Education or adaptive e-learning.

Copyright

© 2021, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Utomo Pujianto
AU  - Harits Ar Rosyid
AU  - Aditya Cahyadi Putra
PY  - 2021
DA  - 2021/10/21
TI  - Performance Comparison of Ensemble-based k-Nearest Neighbor and CART Classifiers for the Classification of Adaptive e-learning User Knowledge Levels
BT  - Proceedings of the 1st UMGESHIC International Seminar on Health, Social Science and Humanities (UMGESHIC-ISHSSH 2020)
PB  - Atlantis Press
SP  - 243
EP  - 251
SN  - 2352-5398
UR  - https://doi.org/10.2991/assehr.k.211020.037
DO  - 10.2991/assehr.k.211020.037
ID  - Pujianto2021
ER  -