K-Nearest Neighbor of Beta Signal Brainwave to Accelerate Detection of Concentration on Student Learning Outcomes

• Main Authors: Saputra, Dimas Chaerul Ekty, Azhari,, Ahmad, Ma’arif, Alfian
• Format: Article PeerReviewed
• Language: English
• Published: International Association of Engineers 2022
• Subjects: Industrial Electronics; Electrical and Electronic Engineering not elsewhere classified; Electrical and Electronic Engineering
• Online Access: https://repository.ugm.ac.id/283279/1/K-NearestNeighborofBetaSignalBrainwavetoAccelerateDetectionofConcentrationonStudentLearningOutcomes.pdf; https://repository.ugm.ac.id/283279/; https://www.engineeringletters.com/
• Institution: Universitas Gadjah Mada

—Intelligence, creativity, emotions, memory, and body movements are human activities controlled by the brain. While humans do an activity, the neural network in the brain produces an electrical current in the form of waves. Brainwaves are one of the biometric features that can be used to identify individual characteristics based on their activity and behavior patterns. Identifying individual characteristics requires a brain activity measurement using an Electroencephalogram (EEG). Measuring brainwaves requires a reliable, prominent, and constant activity stimulation by applying a series of cognitive tasks, such as the Culture Fair Intelligence Test (CFIT) and the Indonesian Competency Test (CT). This research aims to obtain relation patterns and accelerate the detection between brain concentration and learning outcomes. Beta signal acquisition is obtained from junior high school students while performing cognitive tasks. After data is obtained, the signal is extracted using the Fast Fourier Transform (FFT) to get its peak signal. The peak signal from FFT data on CFIT generated an average score of 0.214 with the category of Average. Meanwhile, the peak signal on CT generated an average score of 0.246 with the category "C+". K-Nearest Neighbor (KNN) algorithm is applied to identify patterns from extraction data with K-value=5; then, the accuracy is assessed using K-Fold Cross Validation with K-value=11. The resulting accuracy is 94.59%. Based on the KNN classification results, students' learning outcomes are influenced by their concentration. This research has successfully shortened the CFIT evaluation time from three days to one day. © 2022, International Association of Engineers. All rights reserved.