After model validation, ROC was not show on the performance, but ROC (threshold)

Borriwat

I'm working with only 59 data (7 positve class, 52 negative class). First, I use sample operator to balanced data (undersampling) and create decision tree model with cross-validation. After I check the performance of the model, the ROC showed as the picture. It has only blue line, which is ROC (Threshold).

sgenzer

hi @Borriwat I'm sorry no one has chimed in here. Unfortunately you did not post your data so no one can test. Also if you only have 59 data it is likely you just do not have enough to model anything meaningful here.

Scott

sgenzer

hi @Borriwat I'm sorry no one has chimed in here. Unfortunately you did not post your data so no one can test. Also if you only have 59 data it is likely you just do not have enough to model anything meaningful here.

Scott