Hi, I was wondering how to use 3dsvm on e-rel datasets.
What I was thinking of doing was to use 3dDeconvolve on my dataset and have one regressor per trial. Then, use 3dsvm on the the brik with all the betas, since at that point one could give them labels.
This seems a rather cumbersome method though, involving writing hundreds of time files for the single trials.
Does this make any sense, and is there a more elegant way?
Thanks.
Jack