A super-atomic norm minimization approach to identifying sparse dynamical graphical models

Yin Wang, Mario Sznaier, Octavia I. Camps

Abstract: This paper considers the problem of identifying sparse dynamical graphical models from input/output data. Our main result shows that this problem can be recast into an expanded atomic-norm minimization framework that allows for enforcing block-sparsity. This approach leads to efficient algorithms capable of handling large data sets, unknown inputs and fragmented data records. These results are illustrated with several examples.