IEEE CEC 2005 - Special Session on Complex Adaptive Systems

Complex Adaptive Systems:

Towards Predictive Methodologies

at the 2005 IEEE Congress on Evolutionary Computation
Edinburgh, Scotland - 2-5 September 2005


Session Chairs:

Akira Namatame (
Department of Computer Science, National Defense Academy, 1-10-20 Hashirimizu, Yokosuka, Japan, 239-8686

Chrystopher L. Nehaniv (
Algorithms & Adaptive Systems Research Groups, School of Computer Science, University of Hertfordshire, College Lane, Hatfield, Hertfordshire, United Kingdom AL10 9AB

Scope and Theme:

This special session is concerned with fostering the formation of an active multi-disciplinary community on Complex Adaptive Systems. We especially intend to increase the awareness of researchers in many fields sharing the common view on combining agent-based modeling and the evolutionary computation model in order to develop insight and foster predictive methodologies. Complex adaptive systems involve the study of many agents (constituent components, generally active ones with a simple structures, whose behavior is assumed to follow local rules) and their rich interactions. A basic methodology is to specify how the agents interact, and then observe properties that occur at the collective level in order to discover predictive principles and key descriptive variables for understanding and/or shaping and harnessing the resulting dynamics.

Generally the high-dimensional, non-linear nature of the resulting dynamical systems makes them difficult or impossible to analyze using traditional methods. Agents follow local rules under various constraints (including possibly, e.g. spatial connectivity, geometric, physical, genetic, evolvability, ecological, interactional, sensorimotor, energetic, and information-processing constraints). The resulting dynamics are not necessarily derivable from any principles of analytic calculation. Under the action of evolution, such agents adapt to their environments and other agents' behaviors. The adaptation processes can be massively parallel, depending on the number of agents, and we especially need to explore the relationship between at the individual level and at the collective level. The idea of combining evolutionary computation and agent-based modeling is particularly rich and fresh and applicable to answer these issues.

The emergent phenomena arising from interactions even among a small number of agents and their environment are not well-understood, e.g. in the evolution signaling, communication, and interaction dynamics.

We will invite high quality contributions on a wide variety of topics relevant to the wide research areas of Complex Adaptive Systems. We will especially cover in-depth of important areas such as:

Topics of Interest:

Scientific Program Committee Members

Submissions and Important Dates

Submissions Deadline: 25 April 2005
Notification to Authors: 19 May 2005
Camera-Ready Copies Due: 11 June 2005

All submissions will be peer-reviewed according to IEEE standards. Submissions should be in IEEE two-column format up to 6 pages according to instructions on IEEE CEC website giving format and uploading requirement details. (Authors should indicate this special session when uploading their submission.)

Organized with the support of:
The IEEE Working Group on Artificial Life and Complex Adaptive Systems
The U.K. EPSRC Network on Evolvability in Biological and Software Systems

Special Session Homepage and Updates:

Last update 20 April 2005