1 paper · 1 filter
Rudolf Debelak
The evaluation of machine learning models typically relies mainly on performance metrics based on loss functions, which risk to overlook changes in performance in relevant subgroup…