Most Correlated Arms Identification

Che-Yu Liu, Sébastien Bubeck
Proceedings of The 27th Conference on Learning Theory, PMLR 35:623-637, 2014.

Abstract

We study the problem of finding the most mutually correlated arms among many arms. We show that adaptive arms sampling strategies can have significant advantages over the non-adaptive uniform sampling strategy. Our proposed algorithms rely on a novel correlation estimator. The use of this accurate estimator allows us to get improved results for a wide range of problem instances.

Cite this Paper

BibTeX
@InProceedings{pmlr-v35-liu14, title = {Most Correlated Arms Identification}, author = {Liu, Che-Yu and Bubeck, Sébastien}, booktitle = {Proceedings of The 27th Conference on Learning Theory}, pages = {623--637}, year = {2014}, editor = {Balcan, Maria Florina and Feldman, Vitaly and Szepesvári, Csaba}, volume = {35}, series = {Proceedings of Machine Learning Research}, address = {Barcelona, Spain}, month = {13--15 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v35/liu14.pdf}, url = {https://proceedings.mlr.press/v35/liu14.html}, abstract = {We study the problem of finding the most mutually correlated arms among many arms. We show that adaptive arms sampling strategies can have significant advantages over the non-adaptive uniform sampling strategy. Our proposed algorithms rely on a novel correlation estimator. The use of this accurate estimator allows us to get improved results for a wide range of problem instances.} }
Endnote
%0 Conference Paper %T Most Correlated Arms Identification %A Che-Yu Liu %A Sébastien Bubeck %B Proceedings of The 27th Conference on Learning Theory %C Proceedings of Machine Learning Research %D 2014 %E Maria Florina Balcan %E Vitaly Feldman %E Csaba Szepesvári %F pmlr-v35-liu14 %I PMLR %P 623--637 %U https://proceedings.mlr.press/v35/liu14.html %V 35 %X We study the problem of finding the most mutually correlated arms among many arms. We show that adaptive arms sampling strategies can have significant advantages over the non-adaptive uniform sampling strategy. Our proposed algorithms rely on a novel correlation estimator. The use of this accurate estimator allows us to get improved results for a wide range of problem instances.
RIS
TY - CPAPER TI - Most Correlated Arms Identification AU - Che-Yu Liu AU - Sébastien Bubeck BT - Proceedings of The 27th Conference on Learning Theory DA - 2014/05/29 ED - Maria Florina Balcan ED - Vitaly Feldman ED - Csaba Szepesvári ID - pmlr-v35-liu14 PB - PMLR DP - Proceedings of Machine Learning Research VL - 35 SP - 623 EP - 637 L1 - http://proceedings.mlr.press/v35/liu14.pdf UR - https://proceedings.mlr.press/v35/liu14.html AB - We study the problem of finding the most mutually correlated arms among many arms. We show that adaptive arms sampling strategies can have significant advantages over the non-adaptive uniform sampling strategy. Our proposed algorithms rely on a novel correlation estimator. The use of this accurate estimator allows us to get improved results for a wide range of problem instances. ER -
APA
Liu, C. & Bubeck, S.. (2014). Most Correlated Arms Identification. Proceedings of The 27th Conference on Learning Theory, in Proceedings of Machine Learning Research 35:623-637 Available from https://proceedings.mlr.press/v35/liu14.html.

Related Material