共 3 条
Empowerment as a Generic Utility Function for Agents in a Simple Team Sport Simulation
被引:1
|作者:
Clements, Marcus
[1
]
Polani, Daniel
[1
]
机构:
[1] Univ Hertfordshire, Adapt Syst Res Grp, Sch Comp Sci, Hatfield, Herts, England
来源:
INTERACTIVE COLLABORATIVE ROBOTICS (ICR 2017)
|
2017年
/
10459卷
基金:
欧盟地平线“2020”;
关键词:
Empowerment;
Intrinsic motivation;
Artificial intelligence;
Information theory;
BEHAVIOR;
D O I:
10.1007/978-3-319-66471-2_5
中图分类号:
TP18 [人工智能理论];
学科分类号:
081104 ;
0812 ;
0835 ;
1405 ;
摘要:
Players in team sports cooperate in a coordinated manner to achieve common goals. Automated players in academic and commercial team sports simulations have traditionally been driven by complex externally motivated value functions with heuristics based on knowledge of game tactics and strategy. Empowerment is an information-theoretic measure of an agent's potential to influence its environment, which has been shown to provide a useful intrinsic value function, without the need for external goals and motivation, for agents in single agent models. In this paper we expand on the concept of empowerment to propose the concept of team empowerment as an intrinsic, generic utility function for cooperating agents. We show that agents motivated by team empowerment exhibit recognizable team behaviors in a simple team sports simulation based on Ultimate Frisbee.
引用
收藏
页码:37 / 49
页数:13
相关论文