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


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


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

Computer Science > Machine Learning

Title: Off-policy evaluation for slate recommendation

Abstract: This paper studies the evaluation of policies that recommend an ordered set of items (e.g., a ranking) based on some context---a common scenario in web search, ads, and recommendation. We build on techniques from combinatorial bandits to introduce a new practical estimator that uses logged data to estimate a policy's performance. A thorough empirical evaluation on real-world data reveals that our estimator is accurate in a variety of settings, including as a subroutine in a learning-to-rank task, where it achieves competitive performance. We derive conditions under which our estimator is unbiased---these conditions are weaker than prior heuristics for slate evaluation---and experimentally demonstrate a smaller bias than parametric approaches, even when these conditions are violated. Finally, our theory and experiments also show exponential savings in the amount of required data compared with general unbiased estimators.
Comments: 31 pages (9 main paper, 20 supplementary), 12 figures (2 main paper, 10 supplementary)
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Machine Learning (stat.ML)
Cite as: arXiv:1605.04812 [cs.LG]
  (or arXiv:1605.04812v3 [cs.LG] for this version)

Submission history

From: Adith Swaminathan [view email]
[v1] Mon, 16 May 2016 15:47:21 GMT (2254kb,D)
[v2] Tue, 24 May 2016 10:34:21 GMT (1526kb,D)
[v3] Mon, 6 Nov 2017 22:55:48 GMT (8690kb,D)

Link back to: arXiv, form interface, contact.