driptarc / decision Goto Github PK

Hi, thanks for sharing your great work. Recently, I try to reproduce your experiments on digit 5 with the network used in SHOT (DTNBase). However, it doesn't work (the accuracy is low). Could you please also share the corresponding codes with me?

Hi,
I am interested in studying and reproducing your work. However, as written in the README file, environment.yml file is missing. Could you please check it?
Thank you

Thanks for your impressive work.
I think Eq. (3) in your paper is wrong.
As mentioned in your paper, you want to minimize the conditional entropy term and maximize the entropy of the empirical label distribution. So I think Eq. (3) should be formulated as follows:

By the way, the implementation of your code is the same as above.