Deep Reinforcement Learning in Parameterized Action Space

TL;DR

This paper extends deep reinforcement learning to parameterized action spaces, successfully applying it to RoboCup soccer and achieving more reliable goal scoring than previous champions.

Contribution

It introduces methods for deep RL in structured continuous action spaces, filling a gap in existing research and demonstrating success in a complex simulated environment.

Findings

Agent scores goals more reliably than 2012 RoboCup champion

First successful application of deep RL in parameterized action spaces

Demonstrates potential for complex structured action domains

Abstract

Recent work has shown that deep neural networks are capable of approximating both value functions and policies in reinforcement learning domains featuring continuous state and action spaces. However, to the best of our knowledge no previous work has succeeded at using deep neural networks in structured (parameterized) continuous action spaces. To fill this gap, this paper focuses on learning within the domain of simulated RoboCup soccer, which features a small set of discrete action types, each of which is parameterized with continuous variables. The best learned agent can score goals more reliably than the 2012 RoboCup champion agent. As such, this paper represents a successful extension of deep reinforcement learning to the class of parameterized action space MDPs.