Classification-based Policy Iteration with a Critic
نویسندگان
چکیده
In this paper, we study the effect of adding a value function approximation component (critic) to rollout classification-based policy iteration (RCPI) algorithms. The idea is to use a critic to approximate the return after we truncate the rollout trajectories. This allows us to control the bias and variance of the rollout estimates of the action-value function. Therefore, the introduction of a critic can improve the accuracy of the rollout estimates, and as a result, enhance the performance of the RCPI algorithm. We present a new RCPI algorithm, called direct policy iteration with critic (DPI-Critic), and provide its finite-sample analysis when the critic is based on the LSTD method. We empirically evaluate the performance of DPI-Critic and compare it with DPI and LSPI in two benchmark reinforcement learning problems.
منابع مشابه
Incremental Natural Actor-Critic Algorithms
We present four new reinforcement learning algorithms based on actor-critic and natural-gradient ideas, and provide their convergence proofs. Actor-critic reinforcement learning methods are online approximations to policy iteration in which the value-function parameters are estimated using temporal difference learning and the policy parameters are updated by stochastic gradient descent. Methods...
متن کاملModel-free Adaptive Dynamic Programming for Optimal Control of Discrete-time Affine Nonlinear System ⋆
In this paper, a model-free and effective approach is proposed to solve infinite horizon optimal control problem for affine nonlinear systems based on adaptive dynamic programming technique. The developed approach, referred to as the actor-critic structure, employs two multilayer perceptron neural networks to approximate the state-action value function and the control policy, respectively. It u...
متن کاملAn Analysis of Actor/Critic Algorithms Using Eligibility Traces: Reinforcement Learning with Imperfect Value Function
We present an analysis of actor/critic algorithms, in which the actor updates its policy using eligibility traces of the policy parameters. Most of the theoretical results for eligibility traces have been for only critic's value iteration algorithms. This paper investigates what the actor's eligibility trace does. The results show that the algorithm is an extension of Williams' REINFORCE algori...
متن کاملNatural actor-critic algorithms
We present four new reinforcement learning algorithms based on actor–critic, natural-gradient and function-approximation ideas, and we provide their convergence proofs. Actor–critic reinforcement learning methods are online approximations to policy iteration in which the value-function parameters are estimated using temporal difference learning and the policy parameters are updated by stochasti...
متن کاملData-based approximate policy iteration for nonlinear continuous-time optimal control design
This paper addresses the model-free nonlinear optimal problem with generalized cost functional, and a data-based reinforcement learning technique is developed. It is known that the nonlinear optimal control problem relies on the solution of the Hamilton-Jacobi-Bellman (HJB) equation, which is a nonlinear partial differential equation that is generally impossible to be solved analytically. Even ...
متن کامل