Non-stationary Reinforcement Learning without Prior Knowledge: an Optimal Black-box Approach
Chen-Yu Wei , Haipeng Luo
Session: Paper Awards
Poster: Poster Session 2
Abstract:
We propose a black-box reduction that turns a certain reinforcement learning algorithm with optimal regret in a (near-)stationary environment into another algorithm with optimal dynamic regret in a non-stationary environment, importantly without any prior knowledge on the degree of non-stationarity.
By plugging different algorithms into our black-box,
we provide a list of examples showing that our approach not only recovers recent results for (contextual) multi-armed bandits achieved by very specialized algorithms, but also significantly improves the state of the art for (generalzed) linear bandits, episodic MDPs, and infinite-horizon MDPs in various ways.
Specifically, in most cases our algorithm achieves the optimal dynamic regret $\widetilde{\mathcal{O}}(\min\{\sqrt{LT}, \Delta^{\frac{1}{3}}T^{\frac{2}{3}}\})$ where $T$ is the number of rounds and $L$ and $\Delta$ are the number and amount of changes of the world respectively, while previous works only obtain suboptimal bounds and/or require the knowledge of $L$ and $\Delta$.