Open AccessImproving Accuracy of Classification of Emotions Using EEG Signal and Adaptive PSO
Author(s) -
Payal Ghutke,
Sonali Joshi,
Ruchita Timande*
Publication year - 2021
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/1170/1/012013
Subject(s) - headset , kurtosis , computer science , pattern recognition (psychology) , artificial intelligence , feature (linguistics) , electroencephalography , speech recognition , redundancy (engineering) , signal (programming language) , standard deviation , mathematics , psychology , statistics , telecommunications , linguistics , philosophy , psychiatry , programming language , operating system
For detecting the feelings or emotions of the human being by using brain signals and its classification has been attempt by many researchers. The EEG headset is used for collecting the brain signal of the subject. Because of lots of noise in the input signal taken by EEG headset, various features need to be used as a single feature that cannot give accurate output. The Number of feature used is the key for identifying the emotion of a person automatically. So, we identify various features using an AI based scheme from EEG recorded signals. This various features are saved in the database. Features include mean, maximum, minimum, std. deviation, variance, corr. Coefficient, cov. Coefficient, Median, Kurtosis, Energy, Zero crossing rate. By using Maximum Relevance Minimum Redundancy (MRMR), as per the name we arrange the features to minimum-relevance and maximum-importance of every feature. For removing essential segments PCA is used to diminish the produce feature. The proposed system will outperform and improve the accuracy of emotion detection by using brain wave and Adaptive PSO.