Efficiency of Conformal Testing

Vladimir Vovk, Alexander Gammerman and Glenn Shafer
Additional contact information
Vladimir Vovk: University of London, Royal Holloway
Alexander Gammerman: University of London, Royal Holloway
Glenn Shafer: Rutgers University

Chapter Chapter 9 in Algorithmic Learning in a Random World, 2022, pp 265-304 from Springer

Abstract: Abstract Our question in this chapter is how much we can potentially lose when relying on conformal test martingales as compared with unrestricted testing of either exchangeability or statistical randomness. We will see that at a crude scale we do not lose much when restricting our attention to testing statistical randomness with conformal test martingales. Following the old tradition of the algorithmic theory of randomness, in this chapter we consider the binary case only; this makes our results less relevant in practical applications, but they are still reassuring. We consider two very different settings of the problem. In the first, we are given an arbitrary critical region for testing exchangeability of finite binary sequences of a given length, and the question is whether there exists a conformal test martingale emulating this critical region. In the second, we are given a natural alternative, simple or composite, to the hypothesis of exchangeability, and the question is whether there exists a conformal test martingale competitive with the likelihood ratio of the alternative to the null hypothesis (suitably defined).

Keywords: Kelly gambling; Conformal probability; Exchangeability probability; Randomness probability; Changepoint alternative; Markov alternative (search for similar items in EconPapers)
Date: 2022
References: Add references at CitEc
Citations:

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:sprchp:978-3-031-06649-8_9

Ordering information: This item can be ordered from
http://www.springer.com/9783031066498

DOI: 10.1007/978-3-031-06649-8_9

Access Statistics for this chapter

More chapters in Springer Books from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().