Towards instance-dependent regret optimality in Episodic MDPs with Posterior Sampling
arXiv:2610.26029v1 Announce Type: new Abstract: We study regret minimization in finite-horizon episodic Markov Decision Processes (MDPs). While minimax-optimal algorithms are known, tractable approaches to instance-dependent optimality are still lacking. Motivated by this gap, we introduce $\pi_0$-…
Read the full story at arXiv stat.ML ↗
Timeline · 1 report
- 2026-10-20 04:00 · arXiv stat.ML
Towards instance-dependent regret optimality in Episodic MDPs with Posterior Sampling