Title :
A multivariate time varying autoregressive modeling of nonstationary covariance time series
Author :
Gersch, W. ; Gevins, A. ; Kitagawa, G.
Author_Institution :
University of Hawaii Honolulu, Hawaii
Abstract :
A time varying multivariate auto-regressive modeling of nonstationary time series is shown. An orthogonal polynomial Householder transformation least squares regression analysis modeling and the use of Akaike´s AIC for subset selection are the key ideas in this method. Frequency domain relative power contribution computations yield an interpretation of the changing with time relationships in the analysis of single trial evoked potential electroencephalogram data.
Keywords :
Brain modeling; Electroencephalography; Time varying systems;
Conference_Titel :
Decision and Control, 1983. The 22nd IEEE Conference on
Conference_Location :
San Antonio, TX, USA
DOI :
10.1109/CDC.1983.269585
