Switching regression models and causal inference in the presence of discrete latent variables

zbMATH Open1499.62063arXiv1808.05541MaRDI QIDQ4969082FDOQ4969082

Jonas Peters, Rune Christiansen

Publication date: 5 October 2020

Abstract: Given a response $Y$ and a vector $X=(X^1,dots,X^d)$ of $d$ predictors, we investigate the problem of inferring direct causes of $Y$ among the vector $X$. Models for $Y$ that use all of its causal covariates as predictors enjoy the property of being invariant across different environments or interventional settings. Given data from such environments, this property has been exploited for causal discovery. Here, we extend this inference principle to situations in which some (discrete-valued) direct causes of $Y$ are unobserved. Such cases naturally give rise to switching regression models. We provide sufficient conditions for the existence, consistency and asymptotic normality of the MLE in linear switching regression models with Gaussian noise, and construct a test for the equality of such models. These results allow us to prove that the proposed causal discovery method obtains asymptotic false discovery control under mild conditions. We provide an algorithm, make available code, and test our method on simulated data. It is robust against model violations and outperforms state-of-the-art approaches. We further apply our method to a real data set, where we show that it does not only output causal predictors, but also a process-based clustering of data points, which could be of additional interest to practitioners.

Full work available at URL: https://arxiv.org/abs/1808.05541

Recommendations

• Causal Dantzig: fast inference in linear structural equation models with hidden variables under additive interventions
• scientific article; zbMATH DE number 6276218
• scientific article
• Invariant causal prediction for sequential data
• Identification in causal models with hidden variables

zbMATH Keywords

invariancelatent variableshidden Markov modelscausal discoveryswitching regression models

Mathematics Subject Classification ID

Nonparametric regression and quantile regression (62G08) Nonparametric hypothesis testing (62G10) Causal inference from observational studies (62D20)

Cites Work

• Causation, prediction, and search
• Learning stable and predictive structures in kinetic systems: Benefits of a causal approach
• Causal inference using invariant prediction: identification and confidence intervals
• Title not available (Why is that?)
• Statistics for high-dimensional data. Methods, theory and applications.
• A constrained formulation of maximum-likelihood estimation for normal mixture distributions
• Title not available (Why is that?)
• Title not available (Why is that?)
• Causal Inference for Statistics, Social, and Biomedical Sciences
• Pattern recognition and machine learning.
• Causality. Models, reasoning, and inference
• Title not available (Why is that?)
• Invariant Causal Prediction for Sequential Data
• Extended conditional independence and applications in causal inference
• Estimating high-dimensional intervention effects from observational data
• Characterization and Greedy Learning of Interventional Markov Equivalence Classes of Directed Acyclic Graphs
• Discrete Parameter Variation: Efficient Estimation of a Switching Regression Model
• Mixtures of linear regressions
• Maximum-likelihood estimation for hidden Markov models
• The hidden life of latent variables: Bayesian learning with mixed graph models
• Title not available (Why is that?)
• 10.1162/153244303321897717
• Identifiability of Finite Mixtures
• Asymptotic normality of the maximum-likelihood estimator for general hidden Markov models
• Markov chains: models, algorithms and applications
• Direct Calculation of the Information Matrix via the EM Algorithm
• Estimating the Propagation Rate of a Viral Infection of Potato Plants via Mixtures of Regressions
• A modular system of algorithms for unconstrained minimization
• Asymptotic normality of the maximum likelihood estimator in state space models
• Information-geometric approach to inferring causal directions
• Markov-switching generalized additive models
• Distinguishing cause from effect using observational data: methods and benchmarks
• Title not available (Why is that?)
• Causal inference through a witness protection program

Cited In (1)

• Reconstructing regime-dependent causal relationships from observational time series

Uses Software

• JCI
• CausalKinetiX

This page was built for publication: Switching regression models and causal inference in the presence of discrete latent variables

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4969082)