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

Download:

Current browse context:

cs.LG
< prev | next >
new | recent | 1603

Change to browse by:

cs
math
math.ST
stat
stat.ML

References & Citations

DBLP - CS Bibliography

listing | bibtex

Bookmark

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

Computer Science > Machine Learning

Title: Regret Analysis of the Anytime Optimally Confident UCB Algorithm

Authors: Tor Lattimore
(Submitted on 29 Mar 2016 (v1), last revised 6 May 2016 (this version, v2))
Abstract: I introduce and analyse an anytime version of the Optimally Confident UCB (OCUCB) algorithm designed for minimising the cumulative regret in finite-armed stochastic bandits with subgaussian noise. The new algorithm is simple, intuitive (in hindsight) and comes with the strongest finite-time regret guarantees for a horizon-free algorithm so far. I also show a finite-time lower bound that nearly matches the upper bound.
Comments: 16 pages
Subjects: Machine Learning (cs.LG); Statistics Theory (math.ST); Machine Learning (stat.ML)
Cite as: arXiv:1603.08661 [cs.LG]
  (or arXiv:1603.08661v2 [cs.LG] for this version)

Submission history

From: Tor Lattimore [view email]
[v1] Tue, 29 Mar 2016 07:12:14 GMT (34kb)
[v2] Fri, 6 May 2016 19:06:26 GMT (21kb)
Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Link back to: arXiv, form interface, contact.