论文信息 - Karlsruhe Brainstormers - A Reinforcement Learning Approach to Robotic Soccer

Karlsruhe Brainstormers - A Reinforcement Learning Approach to Robotic Soccer

Our long-term goal is to build a robot soccer team where the decision making part is based completely on Reinforcement Learning (RL) methods. The paper describes the overall approach pursued by the Karlsruhe Brainstormers simulator league team. Main parts of basic decision making are meanwhile solved using RL techniques. On the tactical level, first empirical results are presented for 2 against 2 attack situations.

[1] David Elkind,et al. Learning: An Introduction , 1968 .

[2] Martin L. Puterman,et al. Markov Decision Processes: Discrete Stochastic Dynamic Programming , 1994 .

[3] Andrew G. Barto,et al. Learning to Act Using Real-Time Dynamic Programming , 1995, Artif. Intell..

[4] John N. Tsitsiklis,et al. Neuro-Dynamic Programming , 1996, Encyclopedia of Machine Learning.

[5] J. Filar,et al. Competitive Markov Decision Processes , 1996 .

[6] Craig Boutilier,et al. The Dynamics of Reinforcement Learning in Cooperative Multiagent Systems , 1998, AAAI/IAAI.

[7] Richard S. Sutton,et al. Introduction to Reinforcement Learning , 1998 .

[8] Manuela M. Veloso,et al. Team-Partitioned, Opaque-Transition Reinforced Learning , 1998, RoboCup.

[9] Martin A. Riedmiller. Concepts and Facilities of a Neural Reinforcement Learning Control Architecture for Technical Process Control , 1999, Neural Computing & Applications.

[10] Reinforcement Learning for 3 vs. 2 Keepaway , 2000, RoboCup.

[11] Martin Lauer,et al. An Algorithm for Distributed Reinforcement Learning in Cooperative Multi-Agent Systems , 2000, ICML.