nonlinear transform for P300 BCI

[Brendan]

Hello,

I am writing to ask if anyone has tried adapting BCI2000 so that it can utilize a P300 BCI with a nonlinear pattern recognition approach. This would seem to be a nontrivial feat, as it would involve modifying both operator and the P3 signal processing module.

If anyone is interested in this, please let me know. I ask because I would like to implement an SVM in BCI2000. Recent work (such as Meinecke and Kaper et al's NIPS paper) suggests that SVMs are superior to SWDA in categorizing P300 data.

-Brendan