Mimicking Army Ant Swarm Raids to Search Large Continuous Spaces for Global Optima

Speaker: Garrison W. Greenwood, Ph.D, P.E., Electrical and Computer Engineering, Portland State University, USA.

When: 3:30PM, Wed 8 Nov 2006

Venue: 78-420, GP-South Building

Host: Gordon Wyeth

Abstract:

Science is filled with many challenging problems that require finding global optima on continuous surfaces. Many of these problems are highly complex often exceeding 100 dimensions. Most are provably NP-hard, which means efficient algorithms for finding good solutions don't exist. Nature has already solved many optimization problems of interest using methods and procedures man has only begun to comprehend. New methods are sometimes developed by just mimicking exactly what nature does even if the underlying principles aren't fully understood. In particular, insects have existed for over 150 million years; they are experts at exploring continuous, nonlinear 3-D surfaces for food and suitable habitat. This talk describes a new global optimization algorithm that mimics raiding patterns of army ants (Eciton burchelli) to explore non-convex continuous surfaces for global optima. Preliminary results, over 3-D surfaces, will be presented. An explanation of how this algorithm can be extended to very large N-spaces will be given.

Biography:

Garrison W. Greenwood received the Ph.D. degree in electrical engineering from the University of Washington. After spending more than a decade in industry designing multiprocessor embedded system hardware, he entered academia, where he is now an Associate Professor in the Department of Electrical and Computer Engineering, Portland State University, Portland, OR. His research interests are evolvable hardware and adaptive systems. Dr. Greenwood was a National Science Foundation Scholar-in-Residence at the National Institutes of Health in Washington, DC from 1999 and 2000. He has been an Associate Editor of the IEEE Transactions on Evolutionary Computation since 2000 and currently serves as Vice-President (Conferences) for the IEEE Computational Intelligence Society. He is the co-author of the book Introduction to Evolvable Hardware: A Practical Guide for Designing Self-Adaptive Systems, Wiley-IEEE Press, 2006.