Timezone: »

 
Poster
Dynamic Regret of Policy Optimization in Non-Stationary Environments
Yingjie Fei · Zhuoran Yang · Zhaoran Wang · Qiaomin Xie

Mon Dec 07 09:00 PM -- 11:00 PM (PST) @ Poster Session 0 #162
in Poster Session 1 »

We consider reinforcement learning (RL) in episodic MDPs with adversarial full-information reward feedback and unknown fixed transition kernels. We propose two model-free policy optimization algorithms, POWER and POWER++, and establish guarantees for their dynamic regret. Compared with the classical notion of static regret, dynamic regret is a stronger notion as it explicitly accounts for the non-stationarity of environments. The dynamic regret attained by the proposed algorithms interpolates between different regimes of non-stationarity, and moreover satisfies a notion of adaptive (near-)optimality, in the sense that it matches the (near-)optimal static regret under slow-changing environments. The dynamic regret bound features two components, one arising from exploration, which deals with the uncertainty of transition kernels, and the other arising from adaptation, which deals with non-stationary environments. Specifically, we show that POWER++ improves over POWER on the second component of the dynamic regret by actively adapting to non-stationarity through prediction. To the best of our knowledge, our work is the first dynamic regret analysis of model-free RL algorithms in non-stationary environments.

[ Paper ]

Author Information

Yingjie Fei (Cornell University)
Zhuoran Yang (Princeton)
Zhaoran Wang (Northwestern University)
Qiaomin Xie (Cornell University)

More from the Same Authors