Enhancing neural-network performance via assortativity

Abstract

The performance of attractor neural networks has been shown to depend crucially on the heterogeneity of the underlying topology. We take this analysis a step further by examining the effect of degree-degree correlations -- or assortativity -- on neural-network behavior. We make use of a method recently put forward for studying correlated networks and dynamics thereon, both analytically and computationally, which is independent of how the topology may have evolved. We show how the robustness to noise is greatly enhanced in assortative (positively correlated) neural networks, especially if it is the hub neurons that store the information.

9 pages, 7 figures

References in corpus (15)

• Catastrophic cascade of failures in interdependent networks
• Synchronization in complex networks
• Scale-free brain functional networks
• Sustaining the Internet with Hyperbolic Mapping
• Paths to Synchronization on Complex Networks
• The entropy of network ensembles
• The entropy of randomized network ensembles
• Conservation laws for the voter model in complex networks
• The entropic origin of disassortativity in complex networks
• Zero Pearson Coefficient for Strongly Correlated Growing Trees
• Irregular dynamics in up and down cortical states
• Hub Synchronization in Scale-Free Networks
• Assortative mixing by degree makes a network more unstable
• Redundancy and error resilience in Boolean Networks
• Functional Optimization in Complex Excitable Networks

Cited by in corpus (10)

• Nonnormal amplification in random balanced neuronal networks
• Emergence of assortative mixing between clusters of cultured neurons
• From neurons to epidemics: How trophic coherence affects spreading processes
• Diversity improves performance in excitable networks
• Efficient transmission of subthreshold signals in complex networks of spiking neurons
• Concurrence of form and function in developing networks and its role in synaptic pruning
• The effects of degree distributions in random networks of Type-I neurons
• Average nearest neighbor degrees in scale-free networks
• Growth strategy determines network performance
• Interplay between Network Topology and Dynamics in Neural Systems