<HTML dir=ltr><HEAD>
<META content="text/html; charset=unicode" http-equiv=Content-Type>
<META name=GENERATOR content="MSHTML 8.00.6001.18812"></HEAD>
<BODY>
<DIV dir=ltr id=idOWAReplyText98825>
<DIV dir=ltr>
<DIV><FONT color=#000000 size=2 face=Arial>Hi everyone,</FONT></DIV>
<DIV><FONT size=2 face=Arial></FONT> </DIV>
<DIV><FONT size=2 face=Arial>First, I'd like to express my appreciation for the help I've received lately from members of this listserv. I've posted a number of questions recently, and have received very helpful answers in all cases.</FONT></DIV>
<DIV><FONT size=2 face=Arial></FONT> </DIV>
<DIV><FONT size=2 face=Arial>At the moment, I'm wondering whether anyone has every attempted to implement a Bayesian approach to ICA within EEGLAB specifically. </FONT></DIV>
<DIV><FONT size=2 face=Arial></FONT> </DIV>
<DIV><FONT size=2 face=Arial>I know there are several ICA algorithms written in to EEGLAB, but as far as I can tell, none of them allow for any kind of modelling of priors. Of course, one might argue that integrating prior information undermines 'blind' source separation, but it appears that there is precedent for taking a Bayesian approach to ICA within the digital signal processing field (e.g., Winther & Petersen, 2007, Digital Signal Processing). </FONT></DIV>
<DIV><FONT size=2 face=Arial></FONT> </DIV>
<DIV><FONT size=2 face=Arial>Thanks in advance for your input!</FONT></DIV>
<DIV><FONT size=2 face=Arial>Peter</FONT></DIV>
<DIV><FONT size=2 face=Arial></FONT> </DIV>
<DIV><FONT size=2 face=Arial></FONT> </DIV>
<DIV> </DIV></DIV></DIV>
<DIV dir=ltr id=idSignature16202>
<DIV><FONT color=#000000 size=1 face=Arial>Peter Bachman, PhD</FONT></DIV>
<DIV><FONT size=1 face=Arial>Semel Institute</FONT></DIV>
<DIV><FONT size=1 face=Arial>UCLA Department of Psychiatry & Biobehavioral Sciences</FONT></DIV>
<DIV><FONT size=1 face=Arial><A href="mailto:bachman@psych.ucla.edu">bachman@psych.ucla.edu</A></FONT></DIV></DIV></BODY></HTML>