Timezone: »

 
Adaptive Trust Region Policy Optimization: Convergence and Faster Rates of regularized MDPs
Lior Shani · Yonathan Efroni · Shie Mannor

Sat Dec 14 08:50 AM -- 09:10 AM (PST) @
in The Optimization Foundations of Reinforcement Learning »
Trust region policy optimization (TRPO) is a popular and empirically successful policy search algorithm in Reinforcement Learning (RL) in which a surrogate problem, that restricts consecutive policies to be `close' to one another, is iteratively solved. Nevertheless, TRPO has been considered a heuristic algorithm inspired by Conservative Policy Iteration (CPI). We show that the adaptive scaling mechanism used in TRPO is in fact the natural ``RL version" of traditional trust-region methods from convex analysis. We first analyze TRPO in the planning setting, in which we have access to the model and the entire state space. Then, we consider sample-based TRPO and establish $\tilde O(1/\sqrt{N})$ convergence rate to the global optimum. Importantly, the adaptive scaling mechanism allows us to analyze TRPO in regularized MDPs for which we prove fast rates of $\tilde O(1/N)$, much like results in convex optimization. This is the first result in RL of better rates when regularizing the instantaneous cost or reward.

Author Information

Lior Shani (Technion)
Yonathan Efroni (Technion)
Shie Mannor (Technion)

More from the Same Authors