Evaluation Metrics

Hello,

I am having some doubts on how the evaluation metrics are calculated in ShapeWorks studio. Compactness is quite clear to me, however I don’t understand how specificity and generalizability can be calculated without actively splitting the cohort in train and validation set. Is this done automatically within the Studio?

Thanks in advance for you answer!

Kind regards,

Beatrice

Yes, specificity and generalization are computed internally by using a leave N out approach.