The blessings of multiple causes
From MaRDI portal
Abstract: Causal inference from observational data often assumes "ignorability," that all confounders are observed. This assumption is standard yet untestable. However, many scientific studies involve multiple causes, different variables whose effects are simultaneously of interest. We propose the deconfounder, an algorithm that combines unsupervised machine learning and predictive model checking to perform causal inference in multiple-cause settings. The deconfounder infers a latent variable as a substitute for unobserved confounders and then uses that substitute to perform causal inference. We develop theory for the deconfounder, and show that it requires weaker assumptions than classical causal inference. We analyze its performance in three types of studies: semi-simulated data around smoking and lung cancer, semi-simulated data around genome-wide association studies, and a real dataset about actors and movie revenue. The deconfounder provides a checkable approach to estimating closer-to-truth causal effects.
Recommendations
Cites work
- 10.1162/jmlr.2003.3.4-5.993
- A Generalization of Sampling Without Replacement From a Finite Universe
- A simulation-intensive approach for checking hierarchical models
- Accurate Approximations for Posterior Moments and Marginal Densities
- Automatic differentiation variational inference
- Bayesian checking of the second levels of hierarchical models
- Bayesianly justifiable and relevant frequency calculations for the applied statistician
- Causal inference by using invariant prediction: identification and confidence intervals. With discussion and authors' reply
- Causal inference for statistics, social, and biomedical sciences. An introduction
- Causal Inference Using Potential Outcomes
- Causal Inference With General Treatment Regimes
- Causality. Models, reasoning, and inference
- Characterizing Selection Bias Using Experimental Data
- Comment on J. Neyman and causal inference in experiments and observational studies: On the application of probability theory to agricultural experiments. Essay on principles. Section 9 [Ann. Agric. Sci. 10 (1923), 1--51]
- Confounder detection in high-dimensional linear models using first moments of spectral measures
- Dealing with limited overlap in estimation of average treatment effects
- Double/debiased machine learning for treatment and structural parameters
- Elements of causal inference. Foundations and learning algorithms
- Estimation and Inference of Heterogeneous Treatment Effects using Random Forests
- Estimation of causal effects with multiple treatments: a review and new ideas
- Flexible sensitivity analysis for observational studies without observable implications
- Foundations of Modern Probability
- scientific article; zbMATH DE number 3163305 (Why is no real title available?)
- scientific article; zbMATH DE number 4215168 (Why is no real title available?)
- scientific article; zbMATH DE number 3791440 (Why is no real title available?)
- scientific article; zbMATH DE number 47310 (Why is no real title available?)
- scientific article; zbMATH DE number 1487684 (Why is no real title available?)
- scientific article; zbMATH DE number 194758 (Why is no real title available?)
- scientific article; zbMATH DE number 947414 (Why is no real title available?)
- scientific article; zbMATH DE number 5274806 (Why is no real title available?)
- Identification, inference and sensitivity analysis for causal mediation effects
- Identifying the consequences of dynamic treatment strategies: a decision-theoretic overview
- Learning the parts of objects by non-negative matrix factorization
- Mixed membership stochastic blockmodels
- On the definition of a confounder
- Population structure and cryptic relatedness in genetic association studies
- Probabilistic Principal Component Analysis
- Robust Causal Structure Learning with Some Hidden Variables
- Semi-supervised learning in causal and anticausal settings
- Sensitivity Analysis for the Assessment of Causal Vaccine Effects on Viral Load in HIV Vaccine Trials
- Statistics and Causal Inference
- Structured latent factor analysis for large-scale data: identifiability, estimability, and their implications
- The central role of the propensity score in observational studies for causal effects
- The factorized self-controlled case series method: an approach for estimating the effects of many drugs on many outcomes
- The Probability Approach in Econometrics
- The role of the propensity score in estimating dose-response functions
Cited in
(29)- Doubly debiased Lasso: high-dimensional inference under hidden confounding
- Rejoinder on: ``Hierarchical inference for genome-wide association studies: a view on methodology with software
- Covariate balancing functional propensity score for functional treatments in cross-sectional observational studies
- Convex non-parametric least squares, causal structures and productivity
- Why Are Big Data Matrices Approximately Low Rank?
- Causal inference in genetic trio studies
- Comment on: “The Blessings of Multiple Causes” by Yixin Wang and David M. Blei
- Comment: The challenges of multiple causes
- Comment on “Blessings of Multiple Causes”
- The limits of piecemeal causal inference
- Quantifying direct and indirect effect for longitudinal mediator and survival outcome using joint modeling approach
- The impacts of innovation and trade openness on bank market power: the proposal of a minimum distance cost function approach and a causal structure analysis
- Identifying Effects of Multiple Treatments in the Presence of Unmeasured Confounding
- Dynamic network data envelopment analysis with a sequential structure and behavioural-causal analysis: application to the Chinese banking industry
- Deep latent variable models for generating knockoffs
- Characterization of a robust probabilistic framework for brain magnetic resonance image data distributions
- Nonlinear Causal Discovery with Confounders
- Exploring causal mechanisms and quantifying direct and indirect effects using a joint modeling approach for recurrent and terminal events
- De-confounding Causal Inference Using Latent Multiple-Mediator Pathways
- Inference in High-Dimensional Multivariate Response Regression with Hidden Variables
- Incorporating causal modeling into data envelopment analysis for performance evaluation
- Simultaneous Estimation of Multiple Treatment Effects from Observational Studies
- What is fair? Defining fairness in machine learning for health
- Toward Interpretable Deep Generative Models via Causal Representation Learning
- Mediation analysis with unmeasured confounding between parallel mediators and outcome
- A data-analytics framework for exploring regression associations in multivariate categorical data of firefighters' PTSD
- Identifiability and estimation of causal effects with non-Gaussianity and auxiliary covariates
- Escaping graphical causal priors: multi-generator cooperation for rationalization
- The synthetic instrument: from sparse association to sparse causation
This page was built for publication: The blessings of multiple causes
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5208062)
