scikit learn - Spatial Regularization -
is there way include spatial regularization penalty cost functions in scikit-learn clustering?
more specifically, working neuroscience brain data, every voxels has spatially inherited dependency based on proximity. using 2-classes gaussian mixture learning, obtain, each voxel, probability score of being labeled '1' vs '0' (based on 30-ish samples). task pointless if cannot include regularization based on neighborhood, voxels not independents.
Comments
Post a Comment