On the Identification of Sparse plus Low-rank Graphical Models

This thesis proposes an identification procedure for periodic, Gaussian, stationary reciprocal processes, under the assumption that the conditional dependence relations among the observed variables are mainly due to a limited number of latent variables. The identification procedure combines the sparse plus low-rank decomposition of the inverse covariance matrix of the process and the maximum entropy solution for the block-circulant band extension problem recently proposed in the literature