Evolutionary computation (EC) and multi-agent systems and simulation (MASS) both involve populations of agents. EC is a learning technique by which a population of individual agents adapt according to the selection pressures exerted by an environment; MASS seeks to understand how to coordinate the actions of a population of (possibly selfish) autonomous agents that share an environment so that some outcome is achieved. Both EC and MASS have top-down and bottom-up features. For example, some aspects of multi-agent system engineering (e.g., mechanism design) are concerned with how top-down structure can constrain or influence individual decisions. Similarly, most work in EC is concerned with how to engineer selective pressures to drive the evolution of individual behavior towards some desired goal. Multi-agent simulation (also called agent-based modeling) addresses the bottom-up issue of how collective behavior emerges from individual action. Likewise, the study of evolutionary dynamics within EC (for example in coevolution) often considers how population-level phenomena emerge from individual-level interactions. Thus, at a high level, we may view EC and MASS as examining and utilizing analogous processes. It is therefore natural to consider how knowledge gained within EC may be relevant to MASS, and vice versa; indeed, applications and techniques from one field have often made use of technologies and algorithms from the other field. Studying EC and MASS in combination is warranted and has the potential to contribute to both fields.

The EcoMASS workshop welcomes original submissions on all aspects of Evolutionary Computation and Multi-Agent Systems and Simulation, which include (but are not limited to) the following topics and themes:

• Multi-agent systems and agent-based models utilizing evolutionary computation
• Optimization of multi-agent systems and agent-based models using evolutionary computation
• Evolutionary computation models which rely not on explicit fitness functions but rather implicit fitness functions defined by the relationship to other individuals/agents
• Applications utilizing MASS and EC in combination
• Biological agent-based models (usually called individual-based models) involving evolution
• Evolution of cooperation and altruism
• Genotypic representation of the complex phenotypic strategies of MASS
• Evolutionary learning within MASS (including Baldwinian learning and phenotypic plasticity)
• Emergence and feedbacks
• Open-ended strategy spaces and evolution
• Adaptive individuals within evolving Populations

More information to follow.

ECoMASS will be a half-day workshop, from 8am to 12:10pm on Saturday July 7.

Schedule:

8:00 - 8:10 Introduction to the Workshop
8:10 - 8:35 Evolving Strategies for Games (Sevan G. Ficici)
8:35 - 9:00 Learning and exploiting knowledge in multi-agent task allocation problems (Adam Campbell and Annie S. Wu)
9:00 - 9:25 Agent-Environment Interaction in a Mutli-Agent System: A Formal Model (Gamma Bel-Enguix and M. Dolores Jimenez-Lopez)
9:25 - 9:50 Evolutionary Benefits of Evolvable Component Integration (David Malkin and R. Beau Lotto)

9:50 - 10:20 Coffee Break

10:20 - 10:45 A Framework for Combining Agent-Based Simulation and EC (Bill Rand)
10:45 - 11:10 INVITED TALK: The NERO Game (Kenneth Stanley)
11:10 - 11:35 INVITED TALK: Evolving Multi-Agent Systems (Kagan Tumer)
11:35 - 12:05 Closing Discussion

Workshop: 7 July, 2007.

See GECCO 2007 website for complete details.

Workshop Chairs

• Bill Rand, Northwestern Univeristy
• Sevan G. Ficici, Harvard University

Program Committee

• Ted Belding, NewVectors, LLC
• Michael North, Argonne National Laboratory
• Simon Parsons, Brooklyn College
• Robert G. Reynolds, Wayne State University
• Rick Riolo, University of Michigan
• Moshe Sipper, Ben-Gurion University
• Kagan Tumer, Oregon State University
• Tina Yu, Memorial University of Newfoundland