paper-with-me

홈 › Papers

Learning Adversarial Low-rank Markov Decision Processes with Unknown Transition and Full-information Feedback

2023-11-14 · NeurIPS 2023 11

In this work, we study the low-rank MDPs with adversarially changed losses in the full-information feedback setting. In particular, the unknown transition probability kernel admits a low-rank matrix decomposition \citep{REPUCB22}, and the loss functions may change adversarially but are revealed to the learner at the end of each episode. We propose a policy optimization-based algorithm POLO, and we prove that it attains the $\widetilde{O}(K^{\frac{5}{6}}A^{\frac{1}{2}}d\ln(1+M)/(1-\gamma)^2)$ regret guarantee, where $d$ is rank of the transition kernel (and hence the dimension of the unknown representations), $A$ is the cardinality of the action space, $M$ is the cardinality of the model class, and $\gamma$ is the discounted factor. Notably, our algorithm is oracle-efficient and has a regret guarantee with no dependence on the size of potentially arbitrarily large state space. Furthermore, we also prove an $\Omega(\frac{\gamma^2}{1-\gamma} \sqrt{d A K})$ regret lower bound for this problem, showing that low-rank MDPs are statistically more difficult to learn than linear MDPs in the regret minimization setting. To the best of our knowledge, we present the first algorithm that interleaves representation learning, exploration, and exploitation to achieve the sublinear regret guarantee for RL with nonlinear function approximation and adversarial losses.

📄 PDF Abstract BibTeX arXiv:2311.07876

Code (0)

등록된 구현이 없습니다.

Tasks

Representation Learning

Similar Papers 제목 키워드 기반

Reinforcement Learning in Robust Markov Decision Processes

2013-12-01 · NeurIPS 2013 12 · Shiau Hong Lim, Huan Xu, Shie Mannor

An important challenge in Markov decision processes is to ensure robustness with respect to unexpected or adversarial system behavior while taking advantage of well-behaving parts of the system. We consider a problem set…

reinforcement-learningReinforcement LearningReinforcement Learning (RL)

Learning Adversarial Markov Decision Processes with Bandit Feedback and Unknown Transition

2020-01-01 · ICML 2020 1 · Chi Jin, Tiancheng Jin, Haipeng Luo, Suvrit Sra 외

We consider the task of learning in episodic finite-horizon Markov decision processes with an unknown transition function, bandit feedback, and adversarial losses. We propose an efficient algorithm that achieves O(√L|X|…

Refined Analysis of FPL for Adversarial Markov Decision Processes

2020-08-21 · Yuanhao Wang, Kefan Dong

We consider the adversarial Markov Decision Process (MDP) problem, where the rewards for the MDP can be adversarially chosen, and the transition function can be either known or unknown. In both settings, Follow-the-Pertu…

Learning Adversarial Markov Decision Processes with Delayed Feedback

2020-12-29 · Tal Lancewicki, Aviv Rosenberg, Yishay Mansour

Reinforcement learning typically assumes that the agent observes feedback from the environment immediately, but in many real-world applications (like recommendation systems) the feedback is observed in delay. Thus, we co…

Recommendation Systems

Large Scale Markov Decision Processes with Changing Rewards

2019-05-25 · NeurIPS 2019 12 · Adrian Rivera Cardoso, He Wang, Huan Xu

We consider Markov Decision Processes (MDPs) where the rewards are unknown and may change in an adversarial manner. We provide an algorithm that achieves state-of-the-art regret bound of $O( \sqrt{\tau (\ln|S|+\ln|A|)T}\…