暂无分享,去创建一个
This is a short communication on a Lyapunov function argument for softmax in bandit problems. There are a number of excellent papers coming out using differential equations for policy gradient algorithms in reinforcement learning \cite{agarwal2019optimality,bhandari2019global,mei2020global}. We give a short argument that gives a regret bound for the soft-max ordinary differential equation for bandit problems. We derive a similar result for a different policy gradient algorithm, again for bandit problems. For this second algorithm, it is possible to prove regret bounds in the stochastic case \cite{DW20}. At the end, we summarize some ideas and issues on deriving stochastic regret bounds for policy gradients.
[1] Neil Walton,et al. Regret Analysis of a Markov Policy Gradient Algorithm for Multi-arm Bandits , 2020, ArXiv.
[2] Dale Schuurmans,et al. On the Global Convergence Rates of Softmax Policy Gradient Methods , 2020, ICML.
[3] Sham M. Kakade,et al. Optimality and Approximation with Policy Gradient Methods in Markov Decision Processes , 2019, COLT.
[4] Jalaj Bhandari,et al. Global Optimality Guarantees For Policy Gradient Methods , 2019, ArXiv.