Summary
International Symposium on Nonlinear Theory and its Applications
2008
Session Number:A2L-B
Session:
Number:A2L-B3
On the problem of estimating connectivity from spike recordings in large neuron networks
Federico Bizzarri, Marco Storace, Daniele Stellardo, Oscar De Feo,
pp.-
Publication Date:2008/9/7
Online ISSN:2188-5079
DOI:10.34385/proc.42.A2L-B3
PDF download (138KB)
Summary:
Most of the existing methods to extract information about the interactions within a network of dynamical systems starting from measured data work well for networks with a limited number of interacting units, though they badly scale to networks containing hundreds of elements, the main limiting factor being the computational complexity. This paper deals with a method based on linear regression and particularly conceived for identifying networks of biological neurons. The method complexity scales linearly with the number of network elements. Some examples are proposed in order to validate the method and to evaluate to what extent the quality of the information about the interaction between two neurons is influenced by adding up to one hundred of nodes.