Poster: "Modelling functional Interactions underlying the neural code dynamics"

LISANDRO MONTANGIE; FERNANDO MONTANI

Conferencia; XVII Conference on Nonequilibrium statistical mechanics and nonlinear physics; 2012

Neuronal pairwise interaction models have proven to describe accurately the neurophysiological behavior of the neurons in the retina. However, pairwise approaches do not always provide realistic descriptions of the population dynamics in the cortex as they do not uniquely determine the correlated activity at population level in the cortex. In the framework of a perturbative expansion for short time windows, we formally estimate, up to third order, correlations across a population of neurons under the principle of maximum entropy. This formalism will allow us to investigate within information geometry the role of fundamental constraints in the neural code, and to obtain further insights on the functional connectivity of the network.