Nonlinear Causal Discovery with Confounders

Abstract

This article introduces a causal discovery method to learn nonlinear relationships in a directed acyclic graph with correlated Gaussian errors due to confounding. First, we derive model identifiability under the sublinear growth assumption. Then, we propose a novel method, named the Deconfounded Functional Structure Estimation (DeFuSE), consisting of a deconfounding adjustment to remove the confounding effects and a sequential procedure to estimate the causal order of variables. We implement DeFuSE via feedforward neural networks for scalable computation. Moreover, we establish the consistency of DeFuSE under an assumption called the strong causal minimality. In simulations, DeFuSE compares favorably against state-of-the-art competitors that ignore confounding or nonlinearity. Finally, we demonstrate the utility and effectiveness of the proposed approach with an application to gene regulatory network analysis. The Python implementation is available at https://github.com/chunlinli/defuse.

28 pages, 4 figures, 3 tables

References in corpus (19)

• Adam: A Method for Stochastic Optimization
• High-dimensional graphs and variable selection with the Lasso
• Estimating high-dimensional directed acyclic graphs with the PC-algorithm
• Latent variable graphical model selection via convex optimization
• Deep Neural Networks for Estimation and Inference
• Learning high-dimensional directed acyclic graphs with latent and selection variables
• CAM: Causal additive models, high-dimensional order search and penalized regression
• Identifiability of Gaussian structural equation models with equal error variances
• DAGs with NO TEARS: Continuous Optimization for Structure Learning
• Geometry of the faithfulness assumption in causal inference
• Causal Discovery with Continuous Additive Noise Models
• Deep ReLU network approximation of functions on a manifold
• Identifying confounders using additive noise models
• Penalized Estimation of Directed Acyclic Graphs From Discrete Data
• Causal Discovery with General Non-Linear Relationships Using Non-Linear ICA
• Beware of the Simulated DAG! Causal Discovery Benchmarks May Be Easy To Game
• A Two-Stage Penalized Least Squares Method for Constructing Large Systems of Structural Equations
• Inference for a Large Directed Acyclic Graph with Unspecified Interventions
• The DeCAMFounder: Non-Linear Causal Discovery in the Presence of Hidden Variables

Cited by in corpus (2)

• Towards Human-like Perception: Learning Structural Causal Model in Heterogeneous Graph
• Discovery and inference of a causal network with hidden confounding