1 paper · 1 filter
Ahmed Zaoui, Clément Dombry
Selective prediction, where a model has the option to abstain from making a decision, is crucial for machine learning applications in which mistakes are costly. In this work, we fo…