James Christensen (Committee Member), Chandler Phillips (Advisor), David Reynolds (Committee Member)
Master of Science in Engineering (MSEgr)
Successful classification of human cognitive workload is a vital component in identifying and avoiding potential performance deficits resulting from operator work overload. Previous research suggests that electroencephalogram (EEG) derived features, including center frequency, provide a robust signal which may be used to obtain highly accurate workload classification. The purpose of this work is to investigate evidence of physiological hysteresis and determine if center frequency improves a classifier's ability to correctly identify workload level. Results confirmed that including spectral data creates the most robust feature sets, while center frequency across all bands is equally reliable for classifying workload in the case where cognitive workload level transitions from hard to easy. There is also evidence of physiological signal asymmetry based on transition direction. In summation; spectral features are traditionally most dependable for classifying workload, however center frequency across all bands is an equally viable option for feature representation.
Department or Program
Department of Biomedical, Industrial & Human Factors Engineering
Year Degree Awarded
Copyright 2013, all rights reserved. This open access ETD is published by Wright State University and OhioLINK.