Information Geometry as Applied to Neural Spike Data

Information geometry studies invariant structures of a family of probability distributions. Such a family forms a Riemannian manifold together with a dual pair of affine connections (Amari and Nagaoka 2000). Neural spike train data are of stochastic nature described by probability distributions. Information geometry is applied for elucidating the properties of spiking processes spread over a number of neurons as well as over time.