April 2014
Intermediate to advanced
334 pages
10h 40m
English
Vladimir Vovk, Computer Learning Research Centre, Department of Computer Science, Royal Holloway, University of London, United Kingdom
An appealing property of conformal predictors is their automatic validity under the exchangeability assumption: they make an error with probability not exceeding the prespecified significance level. A major focus of this chapter is on conditional versions of the notion of validity. Other extensions that we consider are a computationally efficient version of conformal prediction and probabilistic prediction. We also discuss classical tolerance regions, which can be regarded as a special case of conformal predictors and a generalization of inductive ...
Read now
Unlock full access