Coordination Developed by Learning from Evaluations

This paper reports on research into the origins of communication and coordination. Several problems with defining communication and coordination are noted. A research methodology is described that circumvents these problems. The methodology is used in an experiment concerning the development of coordination. The aim of the experiment is to see whether a learning agent can use coordination signals, which represent evaluations of its behavior, to learn to coordinate its actions in an unknown environment. The task is a pursuit problem where four agents are needed to capture a randomly moving prey. One of these agents adapts its behavior based on the coordination signals it receives from the three other agents. The development of coordination increased the capture rate in this pursuit problem from an initial 5% to 93%. Thus, in combination with a general learning mechanism, coordination signals may be sufficient for the development of coordination.

[1]  Sarit Kraus,et al.  The role of representation in interaction (abstract): discovering focal points among alternative solutions , 1992, SIGO.

[2]  Paul Vogt,et al.  Perceptual Grounding in Robots , 1997, EWLR.

[3]  Magnus Boman,et al.  Multi-Agent Rationality , 1997, Lecture Notes in Computer Science.

[4]  Luc Steels,et al.  Grounding adaptive language games in robotic agents , 1997 .

[5]  Ezequiel A. Di Paolo,et al.  Social coordination and spatial organization: Steps towards the evolution of communication , 1997 .

[6]  Sarit Kraus,et al.  Coordination without Communication: Experimental Validation of Focal Point Techniques , 1997, ICMAS.

[7]  Sebastian Thrun,et al.  The role of exploration in learning control , 1992 .

[8]  Paul Vogt,et al.  How should a robot discriminate between objects? a comparison between two methods , 1998 .

[9]  C. E. SHANNON,et al.  A mathematical theory of communication , 1948, MOCO.

[10]  P. Marler,et al.  Monkey responses to three different alarm calls: evidence of predator classification and semantic communication. , 1980, Science.

[11]  James L. McClelland,et al.  Parallel distributed processing: explorations in the microstructure of cognition, vol. 1: foundations , 1986 .

[12]  M. Benda,et al.  On Optimal Cooperation of Knowledge Sources , 1985 .

[13]  D. G. Moulton Communication by Chemical Signals , 1968, Science.

[14]  Edwin D. de Jong,et al.  An Accumulative Exploration Method for Reinforcement Learning , 1997 .

[15]  Luc Steels Language learning and language contact , 1997 .

[16]  Donald A. Sofge,et al.  Handbook of Intelligent Control: Neural, Fuzzy, and Adaptive Approaches , 1992 .

[17]  Luc Steels,et al.  The Spontaneous Self-organization of an Adaptive Language , 1995, Machine Intelligence 15.

[18]  C. K. Ogden,et al.  The meaning of meaning: A study of the influence of thought and of the science of symbolism , 1923 .

[19]  Frédéric Kaplan,et al.  A new approach to class formation in multi-agent simulations of language evolution , 1998, Proceedings International Conference on Multi Agent Systems (Cat. No.98EX160).

[20]  Edwin D. de Jong,et al.  Multi-Agent Coordination by Communication of Evaluations , 1997, MAAMAW.