Facial Expression Classification on Student Data Using KAN and MLP: Comparative Study and Ensemble Strategy

DOI

10.2991/978-94-6463-982-7_10How to use a DOI?

Keywords

Emotion recognition; MLP; KAN; ensemble learning; SFER

Abstract

Facial expression classification plays a crucial role in human–computer interaction, particularly in the context of online education, which requires automatic emotion recognition. This study proposes a hybrid approach for facial expression classification by comparing two model architectures: Multilayer Perceptron (MLP) and Kolmogorov–Arnold Network (KAN), and combining them through an ensemble soft voting method. The Student Facial Expression Recognition (SFER) dataset is used to test the model’s performance on seven emotion labels. The experimental results show that MLP excels in recognizing visually dominant expressions such as “surprise”, but has limitations in distinguishing emotions with subtle features. On the other hand, KAN, especially with a polynomial basis, shows improved classification accuracy and balance, with the highest F1-score reaching 0.38 and recall evenly distributed across classes. The ensemble model performs the best overall, with an accuracy of 42% and an F1-score of 0.37, as well as a more balanced prediction distribution. These findings confirm that the combination of linear and nonlinear models can complement each other, providing more robust predictions for facial expression classification in real-world scenarios.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Rina Yulius
AU  - Muchamad Fajri Amirul Nasrullah
AU  - Rajabul Haris
PY  - 2025
DA  - 2025/12/29
TI  - Facial Expression Classification on Student Data Using KAN and MLP: Comparative Study and Ensemble Strategy
BT  - Proceedings of the  8th International Conference on Applied Engineering (ICAE 2025)
PB  - Atlantis Press
SP  - 151
EP  - 172
SN  - 2352-5401
UR  - https://doi.org/10.2991/978-94-6463-982-7_10
DO  - 10.2991/978-94-6463-982-7_10
ID  - Yulius2025
ER  -