Evolving processing speed asymmetries and hemispheric interactions in a neural network model

Authors:
Alexander Grushin;James A. Reggia
Affiliations:
Department of Computer Science, University of Maryland, A.V. Williams Building, College Park, MD 20742, USA;Department of Computer Science, University of Maryland, A.V. Williams Building, College Park, MD 20742, USA
Venue:
Neurocomputing
Year:
2005

Citing 2
Cited 0

Multiobjective evolutionary algorithms: a comparative case studyand the strength Pareto approach

IEEE Transactions on Evolutionary Computation
Cost minimization during simulated evolution of paired neural networks leads to asymmetries and specialization

Cognitive Systems Research

Quantified Score

Hi-index	0.01

Visualization

Abstract

Substantial experimental data suggests that the cerebral hemispheres have different processing speeds, and that this may contribute to hemispheric specialization. Here, we use evolutionary computation models to examine whether asymmetric hemispheric processing speeds and lateralization can emerge in neural networks from the need to respond quickly to stimuli and/or to minimize energy consumption. Simulated neuroevolution produced networks with left-right asymmetric processing speeds whenever fitness depended on energy minimization, but not on quickness of response. The results also provide support for a recent hypothesis that subcortical cross-midline interactions are inhibitory/competitive.