I've performed clustered ensemble empirical mode decomposition on individual subjects and wish to study these at the group level. However, given the slight variation between subjects I am thinking that it will be best to first perform some kind of classification/grouping of the intrinsic mode functions to identify which belong to the same group.
Does anybody have any experience with doing this? There seems to be a debate as to whether or not clustering is sensitive enough and that the Hungarian method might be better. Any advice is most welcomed.
If these are ERPs and EMD signals are from the ERPs. you can correlate each EMD from each patient and create a subjects times subjects matrix. Then you can cluster the matrix.
Thanks Luis. This sounds like a sensible strategy, I will give it a go. If it's feasible I'll share the code to prevent this issue popping up in the future.
- Badges
- Science topic
- Similar topics
- Data Mining
- Classification