Maximum Entropy constraints and higher-order spike correlations

LISANDRO MONTANGIE; FERNANDO MONTANI

Maceio, Brasil

Conferencia; XVIII Conference on Nonequilibrium Statistical Mechanics and Nonlinear Physics; 2014

To understand how information is processed by the brain we need to investigate how information is encoded by the activity of a population of neurons. Although models accounting for pairwise interactions have proved able to capture some of the most important features of population activity at the level of theretina, pairwise models are not enough to provide reliable descriptions of neural systems in general. In this work, we present an information theoretical approach to formally estimate higher-order spike correlations considering the maximum entropy principle. This formalism allows us to investigate maximum entropy costraints for evaluating higherorder spike correlations, within the information geometry framework.