Benchmarking non-conformity score functions in conformal prediction

Sol Erika Boman

Conformal prediction is a useful and versatile alternative to model calibration in machine learning classification. It replaces single-class prediction with prediction sets, guaran…