1
$\begingroup$

I'm using a neural network to do binary classification. My architecture is 500->128->128 followed by a sigmoid cross entropy output layer. Regardless of the number of hidden layers/neurons I use, the average precision over the validation stays fairly constant, at roughly 44%, even though the recall can vary between 5% and 30%. I have tried various alternative loss functions, regularization methods, etc, but the precision remains in this narrow range around 44%.

What could be an explanation for this?

$\endgroup$
  • $\begingroup$ Not sure what the reason for precision staying constant is, but you may be able to increase precision at the expense of recall by varying the threshold. Looking at how precision/recall change as you vary the threshold could be useful for that. $\endgroup$ – rinspy Nov 7 '17 at 12:37
  • $\begingroup$ "Regardless of the number of hidden layers/neurons I use, the average precision over the validation stays fairly constant" - but does the precision vary as you vary the cutoff threshold? $\endgroup$ – rinspy Nov 10 '17 at 8:50

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.