Abstract

Conformal predictors are set predictors that are automatically valid in the sense of having coverage probability equal to or exceeding a given confidence level. Inductive conformal predictors are a computationally efficient version of conformal predictors satisfying the same property of validity. However, inductive conformal predictors have only been known to control unconditional coverage probability. This paper explores various versions of conditional validity and various ways to achieve them using inductive conformal predictors and their modifications. In particular, it discusses a convenient expression of one of the modifications in terms of ROC curves.
Original languageEnglish
Pages (from-to)349 - 376
Number of pages28
JournalMachine Learning
Volume92
Issue number2-3
Early online date8 May 2013
DOIs
Publication statusPublished - 5 Jul 2013

Keywords

  • Inductive conformal predictors
  • Conditional validity
  • Batch mode of learning
  • ROC curves
  • Boosting
  • MART
  • Spam detection

Cite this