Heuristic Q-learning soccer players: A new reinforcement learning approach to RoboCup simulation

Abstract

This paper describes the design and implementation of a 4 player RoboCup Simulation 2D team, which was build by adding Heuristic: Accelerated Reinforcement Learning capabilities to basic players of the well-known UvA Trilearn team. The implemented agents learn by using a recently proposed Heuristic Heinforcement Learning algorithm, the Heuristically Accelerated Q-Learning (HAQL), which allows the use of heuristics to speed up the well-known Reinforcement Learning algorithm Q-Learning. A set of empirical evaluations was conducted in the RoboCup 2D Simulator, and experimental results obtained while playing with other teams shows that the approach adopted hero is very promising. © Springer-Verlag Berlin Heidelberg 2007.