search

slim1017 / VaDE

Python code for paper - Variational Deep Embedding : A Generative Approach to Clustering
288 stars 96 forks source link

The loss function #20

Open goodman1204 opened 4 years ago

goodman1204 commented 4 years ago

I am confused about the loss function in the code, for different parts int the loss functions, it has different + or - symbols between the paper and the code? Can someone help to verify it?

SanGilbert commented 4 years ago

It is actually consistent between this two. In the paper the author gave ELBO form that should be maximized, thus this code turn over the sign to minimize the loss.