Citation

Hossain, Md Sajjad and Biswas, Saroj Kr. and Thounaojam, Dalton Meitei and Khan, Atiya and Siddiquee, Kazy Noor e Alam (2024) Machine Learning Algorithms for Emotion Detection using EEG Signals. In: 2024 IEEE International Conference on Computing, Applications and Systems (COMPAS), 25-26 September 2024, Cox's Bazar, Bangladesh.

Text Machine Learning Algorithms for Emotion Detection using EEG Signals.pdf - Published Version Restricted to Repository staff only Download (224kB)

Abstract

Assessment of clinical subjects’ cognitive functions and state is essential to e-healthcare delivery and developing novel human-machine interfaces. With the advancements in brain physiology research, electroencephalogram (EEG)-based emotion detection has emerged as an intriguing and popular study area. The emotion detection model can be implemented in human mental health monitoring, therapeutic support, law enforcement agencies, driver monitoring, etc. Numerous studies have been conducted in the last few decades to use EEG waves to detect emotions. In these research articles, an exhaustive study was not shown to find the best Machine Learning (ML) algorithms for multi-class emotion detection. Moreover, most of these researchers used a large number of electrodes in the experiment, which creates complexity in real-time setup. Therefore, the primary objective of this study is to identify the most effective ML algorithm that gives the expected accuracy using a single electrode. The proposed research has been done on the SEED-IV dataset to classify four common types of emotions: Happy, Fear, Sad, and Neutral. Different types of standard ML classifiers, including ensemble ML classifiers, have been used in this experiment. The performance of this study has been assessed by splitting and 10-fold cross-validation. Linear Discriminant Analysis (LDA) shows a better classification rate of 72.22% and 72.31% among all the classifiers according to Accuracy and F1 score. This study’s results anticipate opening up new possibilities for neuroscience by demonstrating that single-channel EEG data alone is sufficient for categorizing emotions.

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	Brain-Computer Interfaces (BCIs), Human Machine Interactions
Subjects:	Q Science > Q Science (General) > Q300-390 Cybernetics
Divisions:	Faculty of Engineering (FOE)
Depositing User:	Ms Nurul Iqtiani Ahmad
Date Deposited:	12 Feb 2025 00:36
Last Modified:	12 Feb 2025 00:36
URII:	http://shdl.mmu.edu.my/id/eprint/13414

Downloads

Edit (login required)