Andrea Guidolin: Topological analysis of spike train data

Abstract

The methods of topological data analysis have found in neuroscience a particularly fertile field of application. Their ability to capture and quantify information about shape and connections makes them relevant to study, for example, the geometry and topology of spaces of neural representations of stimuli. I will describe some successful topological methods for the analysis of neural data, focusing on the information they can extract from spike train data, i.e. lists of firing times of different neurons. As a concrete example, I will present results on the spaces of neural responses elicited by particular visual stimuli.