Parametric regret in uncertain Markov decision processes

55 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Safe Policy Improvement by Minimizing Robust Baseline Regret (2016, background): “However, a randomized policy π ( a Randomized policies can do better than their deterministic counterparts, because they allow for hedging among various realizations of the MDP parameters.”
Safe Policy Improvement with an Estimated Baseline Policy (2019, background): “en the dataset, and then ﬁnd the policy for which the minimal performance over the robust MDPs set is maximal. The algorithm however tends to converge to policies that are unnecessarily conservative. Xu and Mannor (2009) considered robust regret over the optimal policy: the algorithm searches for th”
Regret based Robust Solutions for Uncertain Markov Decision Processes (2013, background): “Due to the complexity of computing optimal minimax regret policies [16] , existing algorithms [12] are restricted to handling uncertainty only in reward models and the uncertainties are independent across states.”
Bayesian Robust Optimization for Imitation Learning (2020, background): “We would also like to emphasize that our setting is quite different from robust RL methods which focus on uncertain transition probabilities rather than rewards [46, 57, 59].”