local lyapunov stability

score-aware gradient estimator (1) average-reward reinforcement learning (1) stochastic gradient ascent (1) regret bound (1) stationary distribution (1) exponential family (1) markov decision process (1) policy gradient (1) reinforcement learning (1)