I have a feature called Planning_id with more than 40000 categories. What is the recommended embedding size? I read that:

embedding_dimension = # categories * 0.25 is a good rule of thumb, but I still think that 10000 additional features is a lot.

I am using tensorflow

Any recommendations?

Thank you for your suggestions!


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.