LGPROct 11, 2013

Bandits with Switching Costs: T^{2/3} Regret

arXiv:1310.2997v2110 citations
AI Analysis

This work addresses a fundamental gap in online learning theory, showing for the first time that bandit feedback can lead to worse regret scaling than full-information feedback, with implications for bounded-memory adversaries and adversarial MDPs.

The paper tackles the adversarial multi-armed bandit problem with switching costs, proving that the minimax regret is Θ̃(T^{2/3}), which is significantly higher than the Θ(√T) rate for the full-information version, indicating that bandit feedback can be substantially harder.

We study the adversarial multi-armed bandit problem in a setting where the player incurs a unit cost each time he switches actions. We prove that the player's $T$-round minimax regret in this setting is $\widetildeΘ(T^{2/3})$, thereby closing a fundamental gap in our understanding of learning with bandit feedback. In the corresponding full-information version of the problem, the minimax regret is known to grow at a much slower rate of $Θ(\sqrt{T})$. The difference between these two rates provides the \emph{first} indication that learning with bandit feedback can be significantly harder than learning with full-information feedback (previous results only showed a different dependence on the number of actions, but not on $T$.) In addition to characterizing the inherent difficulty of the multi-armed bandit problem with switching costs, our results also resolve several other open problems in online learning. One direct implication is that learning with bandit feedback against bounded-memory adaptive adversaries has a minimax regret of $\widetildeΘ(T^{2/3})$. Another implication is that the minimax regret of online learning in adversarial Markov decision processes (MDPs) is $\widetildeΘ(T^{2/3})$. The key to all of our results is a new randomized construction of a multi-scale random walk, which is of independent interest and likely to prove useful in additional settings.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes