Identifying confounders using additive noise models

Abstract

We propose a method for inferring the existence of a latent common cause ('confounder') of two observed random variables. The method assumes that the two effects of the confounder are (possibly nonlinear) functions of the confounder plus independent, additive noise. We discuss under which conditions the model is identifiable (up to an arbitrary reparameterization of the confounder) from the joint distribution of the effects. We state and prove a theoretical result that provides evidence for the conjecture that the model is generically identifiable under suitable technical conditions. In addition, we propose a practical method to estimate the confounder from a finite i.i.d. sample of the effects and illustrate that the method works well on both simulated and real-world data.

Appears in Proceedings of the Twenty-Fifth Conference on Uncertainty in Artificial Intelligence (UAI2009)

Cited by in corpus (19)

• CAM: Causal additive models, high-dimensional order search and penalized regression
• Causal Discovery with Continuous Additive Noise Models
• Causal Inference on Discrete Data using Additive Noise Models
• Causality for Machine Learning
• Learning Causal Semantic Representation for Out-of-Distribution Prediction
• Nonlinear Causal Discovery with Confounders
• Detecting low-complexity unobserved causes
• Causal Discovery with Cascade Nonlinear Additive Noise Models
• Distinguishing Cause and Effect via Second Order Exponential Models
• Invariant Gaussian Process Latent Variable Models and Application in Causal Discovery
• Identifying Finite Mixtures of Nonparametric Product Distributions and Causal Inference of Confounders
• Removing systematic errors for exoplanet search via latent causes
• Leveraging directed causal discovery to detect latent common causes
• Causal Hidden Markov Model for Time Series Disease Forecasting
• Conditionally-additive-noise Models for Structure Learning
• Learning Joint Nonlinear Effects from Single-variable Interventions in the Presence of Hidden Confounders
• Causal Effect Identification in Acyclic Directed Mixed Graphs and Gated Models
• Causality and Generalizability: Identifiability and Learning Methods
• Discovery of Causal Additive Models in the Presence of Unobserved Variables