A Hybrid EMD-Wavelet EEG Feature Extraction Method for the Classification of Students' Interest in the Mathematics Classroom. (25th January 2021)

• Record Type: Journal Article
• Authors: Babiker, Areej; Faye, Ibrahima
• Other Names: Aricò Pietro Academic Editor.
• Is Part Of: Computational intelligence and neuroscience. Volume 2021(2021)
• Publication Date: 2021-01-25
• Subjects: Neurosciences -- Data processing -- Periodicals; Computational intelligence -- Periodicals; Computational neuroscience -- Periodicals; 612.80285
• DOI: 10.1155/2021/6617462
• Languages: English
• ISSNs: 1687-5265

Abstract

Abstract : Situational interest (SI) is one of the promising states that can improve student's learning and increase the acquired knowledge. Electroencephalogram- (EEG-) based detection of SI could assist in understanding SI neuroscientific causes that, as a result, could explain the SI role in student's learning. In this study, 26 participants were selected based on questionnaires to participate in the mathematics classroom experiment. SI and personal interest (PI) questionnaires along with knowledge tests were undertaken to measure student's interest and knowledge levels. A hybrid method combining empirical mode decomposition (EMD) and wavelet transform was developed and employed for feature extraction. The proposed method showed significant difference using the multivariate analysis of variance (MANOVA) test and consistently outperformed other methods in the classification performance using weighted k -nearest neighbours (wkNN). The high classification accuracy of 85.7% with the sensitivity of 81.8% and specificity of 90% revealed that brain oscillation patterns of high SI students are somewhat different than students with low or no SI. In addition, the result suggests that the delta rhythm could have a significant effect on cognitive processing.