Analysis of random sequential message passing algorithms for approximate inference

We analyze the dynamics of a random sequential message passing algorithm for approximate inference with large Gaussian latent variable models in a student-teacher scenario. To model nontrivial dependencies between the latent variables, we assume random covariance matrices drawn from rotation invariant ensembles. Moreover, we consider a model mismatching setting, where the teacher model and the one used by the student may be different. By means of dynamical functional approach, we obtain exact dynamical mean-field equations characterizing the dynamics of the inference algorithm. We also derive a range of model parameters for which the sequential algorithm does not converge. The boundary of this parameter range coincides with the de Almeida Thouless (AT) stability condition of the replica-symmetric ansatz for the static probabilistic model.

Automated summary

This paper analyzes a random sequential message passing algorithm for large Gaussian latent variable models. By assuming random covariance matrices and considering model mismatch, the authors obtain dynamical mean-field equations characterizing the dynamics of the inference algorithm, and derive the parameter range for which the algorithm does not converge.