ROC curve frequencies deviate

Asked Apr 02 '23 at 09:05

Active Apr 02 '23 at 09:05

Viewed 9 times

I've got a question of understanding on ROC curve.

For estimation of an optimal cut point for a biomarker, I selected an anchor and performed a ROC analysis. The anchor event is quite rare (about 5%). The event when classifying now with the novel optimal cut point is much more frequent (about 20%). This of course results in a low PPV and high NPV.

Can anyone help me understand this deviation in frequencies. I cannot really wrap my head around it.

Thanks a lot!

search already for some answers, but couldn't really find anything.

asked Apr 02 '23 at 09:05

r_user_7411

Please provide enough code so others can better understand or reproduce the problem. – Community Apr 02 '23 at 21:28
it's not really about the code, but a matter of understanding I'd say – r_user_7411 Apr 18 '23 at 10:06

ROC curve frequencies deviate

0 Answers0