This preview shows page 1. Sign up to view the full content.
Unformatted text preview: UCSB Fall 2010 ECE 594C: Lab 3 (EEG modeling) Assigned: November 1 Report Due: November 10 (electronic submission) Preparation: Get EEG data from Fei, learn how to visualize the brain regions corresponding to a spatial response (already done) As we have seen, the data can be classified into 4 categories, with images corresponding to animal or no animal, and a response for each image that are correct or incorrect. The brain dynamics, and hence spatiotemporal EEG response, for each scenario is expected to be somewhat different, hence we analyze them separately using independent component analysis. (a) Concatenate the readings for several images for each category into a single N T data matrix Y , where N = 31 is the number of electrodes and T = MK is the number of samples, with K denoting the number of images (a parameter to be played with) and M the number of time samples per image. Run PCA on this data, plot the eigenvalue profile, and select the number L of important components (also a parameter to be played with).(also a parameter to be played with)....
View Full Document
- Fall '10