Principal Component Analysis Based on Nonparametric Maximum Entropy #70
Labels
difficulty hard
Hard issue: required deep gensim understanding & high python/cython skills
feature
Issue described a new feature
wishlist
Feature request
Check out the PCA based on preserving entropy (as opposed to reconstruction L2 norm): http://mloss.org/software/view/361/
If it works and is useful, add to gensim.
The text was updated successfully, but these errors were encountered: