Abstract:Birgir Hrafnkelsson, Stefan Siegert, Raphaël Huser, Haakon Bakka, Árni V. Jóhannesson. With modern high-dimensional data, complex statistical models are necessary, requiring computationally feasible inference schemes. We introduce Max-and-Smooth, an approximate Bayesian inference scheme for a flexible class of latent Gaussian models (LGMs) where one or more of the likelihood parameters are modeled by latent additive Gaussian processes. Our proposed inference scheme is a two-step approach. In the first step (Max), the likelihood function is approximated by a Gaussian density with mean and covariance equal to either (a) the maximum likelihood estimate and the inverse observed information, respectively, or (b) the mean and covariance of the normalized likelihood function. In the second step (Smooth), the latent parameters and hyperparameters are inferred and smoothed with the approximated likelihood function. The proposed method ensures that the uncertainty from the first step is correctly propagated to the second step. Because the prior density for the latent parameters is assumed to be Gaussian and the approximated likelihood function is Gaussian, the approximate posterior density of the latent parameters (conditional on the hyperparameters) is also Gaussian, thus facilitating efficient posterior inference in high dimensions. Furthermore, the approximate marginal posterior distribution of the hyperparameters is tractable, and as a result, the hyperparameters can be sampled independently of the latent parameters. We show that the computational cost of Max-and-Smooth is close to being insensitive to the number of independent data replicates, and that it scales well with increased dimension of the latent parameter vector provided that its Gaussian prior density is specified with a sparse precision matrix. In the case of a large number of independent data replicates, sparse precision matrices, and high-dimensional latent vectors, the speedup is substantial in comparison to an MCMC scheme that infers the posterior density from the exact likelihood function. The accuracy of the Gaussian approximation to the likelihood function increases with the number of data replicates per latent model parameter. The proposed inference scheme is demonstrated on one spatially referenced real dataset and on simulated data mimicking spatial, temporal, and spatio-temporal inference problems. Our results show that Max-and-Smooth is accurate and fast.

Maximum likelihood thresholds of Gaussian graphical models and graphical lasso

The Cluster Graphical Lasso for improved estimation of Gaussian graphical models

The Maximum Likelihood Threshold of a Graph

Maximum likelihood thresholds via graph rigidity

Learning Gaussian Networks

Latent variable graphical model selection via convex optimization

Graphical model selection with latent variables

Penalized Maximum Likelihood Estimation of Multi-layered Gaussian Graphical Models

Maximum likelihood thresholds of generic linear concentration models

Two new algorithms for maximum likelihood estimation of sparse covariance matrices with applications to graphical modeling

Learning Latent Variable Gaussian Graphical Models

Adaptive Lasso for High Dimensional Regression and Gaussian Graphical Modeling

Graph Estimation for Matrix-variate Gaussian Data

Bayesian Joint Spike-and-Slab Graphical Lasso

Statistical Inference on High Dimensional Gaussian Graphical Regression Models

Bayesian and Maximum Likelihood Estimation for Gaussian Processes on an Incomplete Lattice

Max-and-Smooth: A Two-Step Approach for Approximate Bayesian Inference in Latent Gaussian Models

L1-Penalized Censored Gaussian Graphical Model

High-Dimensional Gaussian Graphical Regression Models with Covariates

Maximum a posteriori estimation in graphical models using local linear approximation