Preselection Bandits

Viktor Bengs, Eyke Hüllermeier
Proceedings of the 37th International Conference on Machine Learning, PMLR 119:778-787, 2020.

Abstract

In this paper, we introduce the Preselection Bandit problem, in which the learner preselects a subset of arms (choice alternatives) for a user, which then chooses the final arm from this subset. The learner is not aware of the user’s preferences, but can learn them from observed choices. In our concrete setting, we allow these choices to be stochastic and model the user’s actions by means of the Plackett-Luce model. The learner’s main task is to preselect subsets that eventually lead to highly preferred choices. To formalize this goal, we introduce a reasonable notion of regret and derive lower bounds on the expected regret. Moreover, we propose algorithms for which the upper bound on expected regret matches the lower bound up to a logarithmic term of the time horizon.

Cite this Paper


BibTeX
@InProceedings{pmlr-v119-bengs20a, title = {Preselection Bandits}, author = {Bengs, Viktor and H{\"u}llermeier, Eyke}, booktitle = {Proceedings of the 37th International Conference on Machine Learning}, pages = {778--787}, year = {2020}, editor = {Hal Daumé III and Aarti Singh}, volume = {119}, series = {Proceedings of Machine Learning Research}, month = {13--18 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v119/bengs20a/bengs20a.pdf}, url = { http://proceedings.mlr.press/v119/bengs20a.html }, abstract = {In this paper, we introduce the Preselection Bandit problem, in which the learner preselects a subset of arms (choice alternatives) for a user, which then chooses the final arm from this subset. The learner is not aware of the user’s preferences, but can learn them from observed choices. In our concrete setting, we allow these choices to be stochastic and model the user’s actions by means of the Plackett-Luce model. The learner’s main task is to preselect subsets that eventually lead to highly preferred choices. To formalize this goal, we introduce a reasonable notion of regret and derive lower bounds on the expected regret. Moreover, we propose algorithms for which the upper bound on expected regret matches the lower bound up to a logarithmic term of the time horizon.} }
Endnote
%0 Conference Paper %T Preselection Bandits %A Viktor Bengs %A Eyke Hüllermeier %B Proceedings of the 37th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2020 %E Hal Daumé III %E Aarti Singh %F pmlr-v119-bengs20a %I PMLR %P 778--787 %U http://proceedings.mlr.press/v119/bengs20a.html %V 119 %X In this paper, we introduce the Preselection Bandit problem, in which the learner preselects a subset of arms (choice alternatives) for a user, which then chooses the final arm from this subset. The learner is not aware of the user’s preferences, but can learn them from observed choices. In our concrete setting, we allow these choices to be stochastic and model the user’s actions by means of the Plackett-Luce model. The learner’s main task is to preselect subsets that eventually lead to highly preferred choices. To formalize this goal, we introduce a reasonable notion of regret and derive lower bounds on the expected regret. Moreover, we propose algorithms for which the upper bound on expected regret matches the lower bound up to a logarithmic term of the time horizon.
APA
Bengs, V. & Hüllermeier, E.. (2020). Preselection Bandits. Proceedings of the 37th International Conference on Machine Learning, in Proceedings of Machine Learning Research 119:778-787 Available from http://proceedings.mlr.press/v119/bengs20a.html .

Related Material