Causal Inference with Noisy and Missing Covariates via Matrix Factorization

Abstract

Valid causal inference in observational studies often requires controlling for confounders. However, in practice measurements of confounders may be noisy, and can lead to biased estimates of causal effects. We show that we can reduce the bias caused by measurement noise using a large number of noisy measurements of the underlying confounders. We propose the use of matrix factorization to infer the confounders from noisy covariates, a flexible and principled framework that adapts to missing values, accommodates a wide variety of data types, and can augment many causal inference methods. We bound the error for the induced average treatment effect estimator and show it is consistent in a linear regression setting, using Exponential Family Matrix Completion preprocessing. We demonstrate the effectiveness of the proposed procedure in numerical experiments with both synthetic data and real clinical data.

arXiv comment26 pages, 5 figures

Cited by in corpus (9)

• Adapting Text Embeddings for Causal Inference
• Causal Inference in Possibly Nonlinear Factor Models
• Conservative Policy Construction Using Variational Autoencoders for Logged Data with Missing Values
• Off-policy Evaluation in Infinite-Horizon Reinforcement Learning with Latent Confounders
• Intact-VAE: Estimating Treatment Effects under Unobserved Confounding
• Controlling for Unmeasured Confounding in Panel Data Using Minimal Bridge Functions: From Two-Way Fixed Effects to Factor Models
• Imputation and low-rank estimation with Missing Not At Random data
• Active Matrix Factorization for Surveys
• Cause-Effect Deep Information Bottleneck For Systematically Missing Covariates