Online Learning with Feedback Graphs: The True Shape of Regret

Tomáš Kocák, Alexandra Carpentier
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:17260-17282, 2023.

Abstract

Sequential learning with feedback graphs is a natural extension of the multi-armed bandit problem where the problem is equipped with an underlying graph structure that provides additional information - playing an action reveals the losses of all the neighbors of the action. This problem was introduced by Mannor & Shamir (2011) and received considerable attention in recent years. It is generally stated in the literature that the minimax regret rate for this problem is of order $\sqrt{\alpha T}$, where $\alpha$ is the independence number of the graph, and $T$ is the time horizon. However, this is proven only when the number of rounds $T$ is larger than $\alpha^3$, which poses a significant restriction for the usability of this result in large graphs. In this paper, we define a new quantity $R^*$, called the problem complexity, and prove that the minimax regret is proportional to $R^*$ for any graph and time horizon $T$. Introducing an intricate exploration strategy, we define the Exp3-EX algorithm that achieves the minimax optimal regret bound and becomes the first provably optimal algorithm for this setting, even if $T$ is smaller than $\alpha^3$.

Cite this Paper

BibTeX
@InProceedings{pmlr-v202-kocak23a, title = {Online Learning with Feedback Graphs: The True Shape of Regret}, author = {Koc\'{a}k, Tom\'{a}\v{s} and Carpentier, Alexandra}, booktitle = {Proceedings of the 40th International Conference on Machine Learning}, pages = {17260--17282}, year = {2023}, editor = {Krause, Andreas and Brunskill, Emma and Cho, Kyunghyun and Engelhardt, Barbara and Sabato, Sivan and Scarlett, Jonathan}, volume = {202}, series = {Proceedings of Machine Learning Research}, month = {23--29 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v202/kocak23a/kocak23a.pdf}, url = {https://proceedings.mlr.press/v202/kocak23a.html}, abstract = {Sequential learning with feedback graphs is a natural extension of the multi-armed bandit problem where the problem is equipped with an underlying graph structure that provides additional information - playing an action reveals the losses of all the neighbors of the action. This problem was introduced by Mannor & Shamir (2011) and received considerable attention in recent years. It is generally stated in the literature that the minimax regret rate for this problem is of order $\sqrt{\alpha T}$, where $\alpha$ is the independence number of the graph, and $T$ is the time horizon. However, this is proven only when the number of rounds $T$ is larger than $\alpha^3$, which poses a significant restriction for the usability of this result in large graphs. In this paper, we define a new quantity $R^*$, called the problem complexity, and prove that the minimax regret is proportional to $R^*$ for any graph and time horizon $T$. Introducing an intricate exploration strategy, we define the Exp3-EX algorithm that achieves the minimax optimal regret bound and becomes the first provably optimal algorithm for this setting, even if $T$ is smaller than $\alpha^3$.} }
Endnote
%0 Conference Paper %T Online Learning with Feedback Graphs: The True Shape of Regret %A Tomáš Kocák %A Alexandra Carpentier %B Proceedings of the 40th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2023 %E Andreas Krause %E Emma Brunskill %E Kyunghyun Cho %E Barbara Engelhardt %E Sivan Sabato %E Jonathan Scarlett %F pmlr-v202-kocak23a %I PMLR %P 17260--17282 %U https://proceedings.mlr.press/v202/kocak23a.html %V 202 %X Sequential learning with feedback graphs is a natural extension of the multi-armed bandit problem where the problem is equipped with an underlying graph structure that provides additional information - playing an action reveals the losses of all the neighbors of the action. This problem was introduced by Mannor & Shamir (2011) and received considerable attention in recent years. It is generally stated in the literature that the minimax regret rate for this problem is of order $\sqrt{\alpha T}$, where $\alpha$ is the independence number of the graph, and $T$ is the time horizon. However, this is proven only when the number of rounds $T$ is larger than $\alpha^3$, which poses a significant restriction for the usability of this result in large graphs. In this paper, we define a new quantity $R^*$, called the problem complexity, and prove that the minimax regret is proportional to $R^*$ for any graph and time horizon $T$. Introducing an intricate exploration strategy, we define the Exp3-EX algorithm that achieves the minimax optimal regret bound and becomes the first provably optimal algorithm for this setting, even if $T$ is smaller than $\alpha^3$.
APA
Kocák, T. & Carpentier, A.. (2023). Online Learning with Feedback Graphs: The True Shape of Regret. Proceedings of the 40th International Conference on Machine Learning, in Proceedings of Machine Learning Research 202:17260-17282 Available from https://proceedings.mlr.press/v202/kocak23a.html.

Related Material