Open jaschil2 opened 4 years ago
This may not be possible. I have removed the gaussian mixture model as it is slower than minibatch k-means
Active?, Done? if active, add labels, assign and move to backlog or in progress. if done, set release and close
Add progress bar for cluster algorithms. Try to improve performance of GMM by sampling less when creating training set for fitting. Maybe allow users to select sample resolution.