A natural actor-critic algorithm for non-stationary average-reward MDPs achieves dynamic regret O~(sqrt(|S||A|) Delta_T^{1/6} T^{5/6}).
Stochastic multi-armed-bandit problem with non-stationary rewards
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Natural Policy Gradient for Average Reward Non-Stationary RL
A natural actor-critic algorithm for non-stationary average-reward MDPs achieves dynamic regret O~(sqrt(|S||A|) Delta_T^{1/6} T^{5/6}).