[Eeglablist] Questions about Kalman Filter

[Ibtissem KHOUAJA BENFRADJ]

Dear list,
 I need your help to overcome some problems that I met.
I read the article "Kalman filter parameters as a new feature vector for EEG BCI applications." (Amir Omidvarnia)
I also tested the algorithm DEKF on Matlab on a portion of the EEG signal(100 points, order 10) to generate the autoregressive matrix. As a resulted, I have got a matrix [1 * 10] for each cahnnel.

My questions are:
knowing that the EEG signal is a non-periodic and non-stationary signal, how can we apply there the Kalman filter?
>From the AR matrix, how can we regain the starting signal (I used  'filter' function under matlab but it does not give the desired result)?
Finally, do you have an idea about the decomposition of the EEG signal in Fourier series?
Thank you in advance for your answers.
cordially,
--------------------------------------------------------------------------------------------------------------------Ibtissem KHOUAJA BENFRADJPhD in computer scienceSpeciality Signal ProcessingLaboratory LTIM, University of Monastir, Tunisiahttp://www.labtim.org/accueil.phpLaboratory LIGM, Univerisity of Paris-East, France http://ligm.u-pem.fr/

 		 	   		   		 	   		   		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20150323/7bb26dd9/attachment.html>