I am using the TensorFlow to do a simple linear classification using logistic regression. The graph included from the TensorBoard displays what they call the fraction of zero weights. How do I interpret this in terms of model evaluation? I am assuming this is good since I got the good results in terms of loss, precision, recall, etc but not sure.

Thank you.

  • $\begingroup$ Set your smoothing to zero. The fraction is 0 at step 100, which appears to be the first real time it is sampled $\endgroup$ – kbrose Sep 26 '19 at 23:47

If you are performing linear (logistic) regression your weights are simply your $\beta_i$. If none of them are $0$ that simply means all features are 'important' to some degree.

| improve this answer | |

Your Answer

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

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