An Assessment of Linear and Nonlinear Features for Detecting Olfactory Stimulus in EEG

The sense of smell is one of the oldest senses of humankind and is able to provide valuable information from the mood of a person to purchase intention. In this study, five non-linear features; 3 Hjorth Parameters namely, activity, complexity, and mobility, Higuchi's Fractal Dimension, and Lemp...

Full description

Saved in:
Bibliographic Details
Main Authors: Akbugday, Burak, Akan, Aydin, Pehlivan, Sude, Sadighzadeh, Reza
Format: Conference Proceeding
Language:English
Subjects:
electroencephalogram (EEG)
Electroencephalography
Feature extraction
Fractals
Higuchi fractal dimension
Hjorth parameters
Lempel-Ziv complexity
machine learning
Mood
neuromarketing
Olfactory
olfactory stimulus
Protocols
Support vector machines
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The sense of smell is one of the oldest senses of humankind and is able to provide valuable information from the mood of a person to purchase intention. In this study, five non-linear features; 3 Hjorth Parameters namely, activity, complexity, and mobility, Higuchi's Fractal Dimension, and Lempel-Ziv Complexity were used to differentiate EEG signals of participants with or without being subjected to olfactory stimuli using several machine learning methods. Experimental results were compared to our previous study where classification was performed using EEG sub-band powers. It was concluded that non-linear features were superior in differentiating olfactory stimuli, especially for frontal, temporal, and occipital channels.
ISSN:2687-7783
DOI:10.1109/TIPTEKNO56568.2022.9960190