From: Jim P. <ji...@ua...> - 2003-01-31 21:48:27
|
John C Atkeson wrote: > Thanks for the links on principle component analysis, I will now > devote some effort trying to understand what it is. ;-) Well, to give you a start, the way it was sold to me was as a black box into which you put a large number of recordings (of brainwaves mixed with evoked potentials in this case), and out of the other side comes a list of wave-shapes which are statistically the most significant ones that the recordings can be broken down into. So, in theory, it would likely pull out the evoked potential shape as one of these components. Also, I think it was claimed that you could use the matrixes built up from the analysis to detect one of these components at the moment it occurs. (Maybe this is as simple as convolving the evoked potential shape with the incoming signal; I never did get 100% understanding of this.) I hope I am not side-tracking you with all this PCA stuff, though. Jim P.S. If you are completely stuck for EEG recordings, you could look at the example ones on the BWView page on my site (provided by Jim Meissner; there are more on his site). There is occasional sync loss to watch out for in these files. -- Jim Peters (_)/=\~/_(_) ji...@ua... (_) /=\ ~/_ (_) Uazú (_) /=\ ~/_ (_) http:// B'ham, UK (_) ____ /=\ ____ ~/_ ____ (_) uazu.net JUST SAY NO! to corrupt audio CDs. For details: http://uazu.net/CD |