Role of Chi-Square in Enhancing the Accuracy of Classifiers in Emotion Recognition

DOI

10.2991/978-94-6463-716-8_41How to use a DOI?

Keywords

Emotion Recognition; Action Units; Chi-Square; K-Nearest Neighbour; Random Forest; Support Vector Machine; Logistic Regression; Naïve Bayes

Abstract

Emotion recognition has been a fast-emerging field in artificial intelligence. It has applications in the healthcare sector, human-computer interaction, and behavioral studies. Feature selection is vital in improving the accuracy and efficiency of classifiers in this field, especially for high-dimensional data. This paper investigates the utilization of the chi-square test for improving the performance of the various classifiers for emotion recognition. The experiment uses the benchmarking CK+ (Cohn-Kanade) database that contains labeled facial expression images. The chi-square test is used to select features in the most relevant manner corresponding to Action Units (AU), related to emotions, targeting here “sadness”. Reduced dimensionality of the dataset gives rise to the chi-square test while removing irrelevant features and alleviating noise and computational complexity associated with it. Accuracy obtained from the K-Nearest Neighbour (k-NN) classifier is compared before as well as after applying chi-square feature selection. The preliminary results show high improvement in the classification accuracy score, 79% becoming 87%, which indicates high effectiveness in refining the features by the application of chi-square. Thus, this experiment shows how statistical feature selection techniques have relevance in optimizing various machine learning classifiers for accurate emotion recognition. After observing the improvement in the performance of KNN, the performance of classifiers like Random Forest (RF), Support Vector Machine (SVM), Logistic Regression (LR) and Naïve Bayes (NB) was also checked. Improvement in F1-score for each classifier was recorded after applying chi-square statistical tool on the relevant extracted features from the images. The possible combination of the chi-square attribute with classifiers can really lead to robust and highly efficient systems in emotion recognition technology, leading to more high-end applications of AI-powered systems.

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  - Pushpa Pathak
PY  - 2025
DA  - 2025/05/26
TI  - Role of Chi-Square in Enhancing the Accuracy of Classifiers in Emotion Recognition
BT  - Proceedings of the International Conference on Recent Advancements and Modernisations in Sustainable Intelligent Technologies and Applications (RAMSITA 2025)
PB  - Atlantis Press
SP  - 526
EP  - 538
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6463-716-8_41
DO  - 10.2991/978-94-6463-716-8_41
ID  - Pathak2025
ER  -