Multiple Model-Based Reinforcement Learning

TLDR

The paper proposes a modular reinforcement learning architecture, Multiple Model‑Based Reinforcement Learning (MMRL), for nonlinear, nonstationary control tasks. MMRL decomposes tasks into spatial‑temporal domains, employing modules that combine state‑prediction models with RL controllers, weighted by a softmax responsibility signal derived from prediction errors, and is formulated for both discrete‑time finite‑state and continuous‑time continuous‑state settings. MMRL achieved successful performance on a nonstationary hunting task in a grid world and on a nonlinear, nonstationary pendulum swing‑up task with varying parameters.

Abstract

We propose a modular reinforcement learning architecture for nonlinear, nonstationary control tasks, which we call multiple model-based reinforcement learning (MMRL). The basic idea is to decompose a complex task into multiple domains in space and time based on the predictability of the environmental dynamics. The system is composed of multiple modules, each of which consists of a state prediction model and a reinforcement learning controller. The "responsibility signal," which is given by the softmax function of the prediction errors, is used to weight the outputs of multiple modules, as well as to gate the learning of the prediction models and the reinforcement learning controllers. We formulate MMRL for both discrete-time, finite-state case and continuous-time, continuous-state case. The performance of MMRL was demonstrated for discrete case in a nonstationary hunting task in a grid world and for continuous case in a nonlinear, nonstationary control task of swinging up a pendulum with variable physical parameters.