Selecting informative conformal prediction sets with false coverage rate control

arXiv:2403.12295v1 Announce Type: cross
Abstract: In supervised learning, including regression and classification, conformal methods provide prediction sets for the outcome/label with finite sample coverage for any machine learning predictors. We consider here the case where such prediction sets come after a selection process. The selection process requires that the selected prediction sets be `informative' in a well defined sense. We consider both the classification and regression settings where the analyst may consider as informative only the sample with prediction label …

abstract arxiv case classification control coverage false machine machine learning math.st prediction process rate regression sample stat.ml stat.th supervised learning type