We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

math.PR

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Mathematics > Probability

Title: Computable randomness is about more than probabilities

Abstract: We introduce a notion of computable randomness for infinite sequences that generalises the classical version in two important ways. First, our definition of computable randomness is associated with imprecise probability models, in the sense that we consider lower expectations (or sets of probabilities) instead of classical 'precise' probabilities. Secondly, instead of binary sequences, we consider sequences whose elements take values in some finite sample space. Interestingly, we find that every sequence is computably random with respect to at least one lower expectation, and that lower expectations that are more informative have fewer computably random sequences. This leads to the intriguing question whether every sequence is computably random with respect to a unique most informative lower expectation. We study this question in some detail and provide a partial answer.
Subjects: Probability (math.PR); Logic (math.LO)
Cite as: arXiv:2005.00471 [math.PR]
  (or arXiv:2005.00471v3 [math.PR] for this version)

Submission history

From: Floris Persiau [view email]
[v1] Fri, 1 May 2020 16:24:06 GMT (56kb)
[v2] Fri, 5 Jun 2020 16:58:07 GMT (51kb)
[v3] Tue, 22 Sep 2020 13:44:34 GMT (52kb)

Link back to: arXiv, form interface, contact.