Title: Learning directed acyclic graphs with hidden variables via latent Gaussian graphical model selection
Authors: Benjamin Frot - ETH Zürich (Switzerland) [presenting]
Preetam Nandy - University of Pennsylvania (United States)
Marloes Maathuis - ETH Zurich (Switzerland)
Abstract: A new method is introduced to estimate the Markov equivalence class of a directed acyclic graph (DAG) in the presence of hidden variables, in settings where the underlying DAG among the observed variables is sparse, and there are a few hidden variables that have a direct effect on many of the observed ones. Building on the so-called low-rank plus sparse framework, we suggest a two-stage approach which first removes unwanted variation using latent Gaussian graphical model selection, and then estimates the Markov equivalence class of the underlying DAG by applying GES. This approach is consistent in certain high-dimensional regimes and performs favourably when compared to the state of the art, both in terms of graphical structure recovery and total causal effect estimation.