Let's say that I have a model that can recognize Cats and Dogs. However, when I use a picture of a Cup or Human it generates a random prediction at 0.70 confidence. Should I use sigmoid instead of softmax?

Or is there any fix for preventing massive confidence on new class object?

  • $\begingroup$ For binary classification sigmoid is softmax. So it shouldn’t matter much. $\endgroup$ Jun 6, 2019 at 8:07

1 Answer 1


I haven't actually played around with this problem, but here are a few resources:

Edit: I just found the following questions here on DS.se:


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.