model · first seen 8 Oct 2026
adversarial Markov decision process
Also: AMDP, adversarial MDP
A decision process in which loss functions may be chosen adversarially, with regret measured against the best fixed policy.
Mentions per week
PapersCodeDiscussion
Earliest sources
- Toward Optimal Regret in Adversarial MDPs with Stochastic Hard ConstraintsPapers · 8 Oct 2026 · Qian Zuo, Francesco Emanuele Stradi
- Introducing Feedback Delays to Non-Delayed Adversarial MDPs: Optimal $\sqrt S$ and $\sqrt K$ Dependencies in Tabular and Linear AMDPsPapers · 12 Nov 2026 · Aarush Kulkarni, Khang Nguyen, William Chang, Yan Dai
- Improved Regret via Variance-Aware Policy Optimization in Adversarial MDPsPapers · 12 Nov 2026 · Ofir Schlisselberg, Alon Cohen, Yishay Mansour
Latest sources
- Improved Regret via Variance-Aware Policy Optimization in Adversarial MDPsPapers · 12 Nov 2026 · Ofir Schlisselberg, Alon Cohen, Yishay Mansour
- Introducing Feedback Delays to Non-Delayed Adversarial MDPs: Optimal $\sqrt S$ and $\sqrt K$ Dependencies in Tabular and Linear AMDPsPapers · 12 Nov 2026 · Aarush Kulkarni, Khang Nguyen, William Chang, Yan Dai
- Toward Optimal Regret in Adversarial MDPs with Stochastic Hard ConstraintsPapers · 8 Oct 2026 · Qian Zuo, Francesco Emanuele Stradi