Gaussian Graphical Models for Partially Observed Multivariate Functional Data

TL;DR

This paper introduces an EM algorithm for inferring Gaussian graphical models from partially observed multivariate functional data, addressing challenges in statistical dependence estimation when data are incomplete.

Contribution

It develops a novel EM-based penalized inference method for functional graphical models under partial observability and Gaussian assumptions.

Findings

The method effectively estimates dependencies in simulated data.

Application to electricity market data demonstrates practical utility.

Results show improved inference with partial observations.

Abstract

In many applications, the variables that characterize a stochastic system are measured along a second dimension, such as time. This results in multivariate functional data and the interest is in describing the statistical dependences among these variables. It is often the case that the functional data are only partially observed. This creates additional challenges to statistical inference, since the functional principal component scores, which capture all the information from these data, cannot be computed. Under an assumption of Gaussianity and of partial separability of the covariance operator, we develop an EM-type algorithm for penalized inference of a functional graphical model from multivariate functional data which are only partially observed. A simulation study and an illustration on German electricity market data show the potential of the proposed method.