New Algorithm Achieves Improved Regret Bounds for Adversarial MDPs
Researchers report a variance-aware policy optimization method for episodic adversarial MDPs with unknown dynamics and full-information rewards, achieving regret with leading term O~(sqrt(H^3 S A K)).
Read the full story at arXiv cs.LG ↗
Timeline · 1 report
- 2026-11-16 05:00 · arXiv cs.LG
Improved Regret via Variance-Aware Policy Optimization in Adversarial MDPs