Separability, stability and persistence of potential representation schemes
The manner by which objects in the external word are represented in neuronal networks is still not understood, yet it is widely believed that such representations involve groups of neurons activated in particular spatial and/or temporal orders. This project seeks to measure the separability of potential neural representation schemes by employing a set of classification approaches, characterizing the relations between multi-electrode array (MEA) topologies and the ability to separate activity patterns invoked from different sources, and examine the invariance of such classification schemes to sampling regimes and the passage of time.