Call for papers

We think that three new kinds of challenges are more and more important in Exploration and Exploitation :

  • Contextual bandit problems with large number of contexts and (unknown) correlations between arms with respect to the context,
  • Numerical difficulties due to large scale problems, with the need for very fast, robust algorithms able to learn and reply in a few milliseconds per query on a standard computer,
  • New theoretical breakthroughs in the analysis of bandit algorithms.

Balancing exploration and exploitation is of high practical importance in large-scale applications on the web and also in information retrieval, where it is desired to help users quickly access the information they are looking for. This requires learning what users are interested in (exploration) and, based on these learnings, to serve adequate targeted content (exploitation).

The workshop will be single-day, made  of invited talks and presentations of contributed work, with time for discussion. Depending on the quality and compatibility with the workshop objectives, slots for brief talks and posters will be allocated.

We invite contributions on the topic of exploration and exploitation in various domains including (but not limited to) bandit games, online optimisation, reinforcement learning, tree search, recommender systems, information retrieval, etc. Topics of interest include: applications and practical performance of techniques to trade exploration and exploitation, benchmark/comparison of different techniques, computational challenges in large scale applications, new techniques, theoretical analyses, etc.

Contributions should be communicated to the programme committee (the organisers) by way of an extended abstract (a pdf file, from 4 to 8 pages formatted in the ICML conference paper style), sent by email to jeremie.mary[at] before May 7th 2012, midnight WST.

Dual Submissions Policy: we accept contributions already published if theses conditions are filled :

  • First publication of the work in the previous year
  • If the work is also an ICML’12 paper then the workshop paper must emphases the link between the paper and the challenge.

You should send us your best papers !!!

You can find the last year proceedings on  JMLR Workshop series