I am trying to choose the best threshold for a binary classification problem that maximises F1 measure. Currently, I am manually analysing the F1 measure at thresholds 0.1-0.9 in steps of 0.1. I am also plotting ROC curve using tpr and fpr. However, is there any approach to automatically select a threshold that separates the maximum number of positives examples from negative ones (or maximizes F1 measure)?


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.