How can I make a classifier that adapts to an increasing number of classes in the dataset.

  • How many nodes should I use as output?
  • If I use a softmax I need to specify the classes but if the classes increase, how can I still use part of the pre-trained architecture and make it learn more classes?
  • Is adding nodes or using redundancy in the output layer the best choices?
  • $\begingroup$ Is there a max number of classes? $\endgroup$ – Fernando Oct 6 '17 at 20:28
  • $\begingroup$ Yes. But if the number of classes is very big, should I just use a classifier with the initial number of nodes equal to the max value of classes? $\endgroup$ – A.Rios Oct 8 '17 at 1:48
  • $\begingroup$ I don't see other way around. $\endgroup$ – Fernando Oct 9 '17 at 13:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.