Hi,
I want to extract EEG features like CSP, mean, skewness etc., from 118 channels. I want to know that shall I take CSP feature of 1st channel as one of the features, and CSP feature of 2nd channel as another feature?? or CSP feature for all 118 channel should be taken as a single feature (i.e., no of trials x one column).
Thanks
Thenmozhi
Ciarán McGeady
In addition to being used for feature extraction, the CSP algorithm is also a feature reduction techinique. Use all channels to builid a CSP projection matrix. This first and last m rows of this matrix can be used to spatially filter other trials. This will reduce the number of channels. Take the logarithmic variance of these spatially filtered trials. These are your features.
0 votes
0 thanks
- Badges
- Science topic
- Similar topics
- Journal Impact Factor
More Thenmozhi Thangaraj's questions See All
Similar questions and discussions