Colloquium Details
A Neural-Group Basis for Evolving and Developing Neural Networks
| Author: | Keith Downing Norwegian University of Science and Technology |
|---|---|
| Date: | June 20, 2008 |
| Time: | 13:30 |
| Location: | 220 Deschutes |
Abstract
This work investigates an intermediate abstraction level, that of neural groups, for modelling the development of complex artificial neural networks. Based on Neural Darwinism (Edelman, 2000}, Displacement Theory (Deacon, 1998) and The Neuromeric Mode (Striedter, 2005}, our DEACANN system avoids the complexities of axonal and dendritic growth while maintaining key aspects of cell signalling, competition and cooperation that appear to govern the formation of neural topologies in nature. DEACANN also includes a genetic-algorithm for evolving developmental recipes, and the mature networks can employ several forms of learning.
This presentation, based on a paper presented at the Genetic and Evolutionary Computation Conference (GECCO) in 2007, will serve as the motivation for one or more informal follow-up lectures on the issues involved in evolving tree- and graph-like structures (such as neural networks).
Biography
Dr. Keith Downing has been Professor of Computer Science at The Norwegian University of Science and Technology (Trondheim, Norway) since 1996. He received his B.S. in Mathematics from Bucknell University in 1983, and his Ph.D. in Computer Science (Artificial Intelligence) from the University of Oregon in 1990. Since then he has worked as a Post-doc in Linkoping, Sweden, and building environmental simulators at SINTEF in Trondheim, Norway.
Dr. Downing's primary research interests include: Artificial Intelligence, Artificial Life, Evolutionary Computation, Neural Networks, and Computational Neuroscience.