The primary visual cortex of the mouse is an ideal model system in which we can study network restructuring during learning, due to the well-described functional properties of its neurons, the existence of a robust learning paradigm, and the feasibility of recording large populations of neurons. Here we aim to identify functional groups of cortical neurons based on the robust estimation of their correlations and observe the topological re-structuring of the network they form. We will build comprehensive Bayesian multi-level models that respect the multidimensional structure of the data for estimation and interpretation of our results. We hypothesize that we will observe major restructuring of functional cortical sub-networks upon learning.