Reinforcement learning algorithm provably optimal for logistic decision processes

Original: Minimax Optimal Variance-Aware Regret Bounds for Multinomial Logistic MDPs

Source: arxiv.org ↗