UvA-DARE (Digital Academic Repository) Foundations of structural causal models with cycles and latent variables

Structural causal models (SCMs), also known as (nonparametric) structural equation models (SEMs), are widely used for causal modeling purposes. In particular, acyclic SCMs, also known as recursive SEMs, form a well-studied subclass of SCMs that generalize causal Bayesian networks to allow for latent confounders. In this paper, we investigate SCMs in a more general setting, allowing for the presence of both latent confounders and cycles. We show that in the presence of cycles, many of the convenient properties of acyclic SCMs do not hold in general: they do not always have a solution; they do not always induce unique observational, interventional and counterfactual distributions; a marginalization does not always exist, and if it exists the marginal model does not always respect the latent projection; they do not always satisfy a Markov property; and their graphs are not always consistent with their causal semantics. We prove that for SCMs in general each of these properties does hold under certain solvability conditions. Our work generalizes results for SCMs with cycles that were only known for certain special cases so far. We introduce the class of simple SCMs that extends the class of acyclic SCMs to the cyclic setting, while preserving many of the convenient properties of acyclic SCMs. With this paper, we aim to provide the foundations for a general theory of statistical causal modeling with SCMs.

[1]  Joris M. Mooij,et al.  Causal Calculus in the Presence of Cycles, Latent Confounders and Selection Bias , 2019, UAI.

[2]  Joseph Y. Halpern,et al.  Abstracting Causal Models , 2018, AAAI.

[3]  Stefan Bauer,et al.  Learning stable and predictive structures in kinetic systems , 2018, Proceedings of the National Academy of Sciences.

[4]  Mélanie Frappier,et al.  The Book of Why: The New Science of Cause and Effect , 2018, Science.

[5]  Joris M. Mooij,et al.  Constraint-based Causal Discovery for Non-Linear Structural Causal Models with Cycles and Latent Confounders , 2018, UAI.

[6]  Joris M. Mooij,et al.  Beyond Structural Causal Models: Causal Constraints Models , 2018, UAI.

[7]  J. Mooij,et al.  Causal Modeling of Dynamical Systems , 2018, 1803.08784.

[8]  Soren Wengel Mogensen,et al.  Markov equivalence of marginalized local independence graphs , 2018, The Annals of Statistics.

[9]  Bernhard Schölkopf,et al.  Elements of Causal Inference: Foundations and Learning Algorithms , 2017 .

[10]  J. Mooij,et al.  Markov Properties for Graphical Models with Cycles and Latent Variables , 2017, 1710.08775.

[11]  Bernhard Schölkopf,et al.  Causal Consistency of Structural Equation Models , 2017, UAI.

[12]  Joris M. Mooij,et al.  Joint Causal Inference from Multiple Contexts , 2016, J. Mach. Learn. Res..

[13]  R. Evans.  Margins of discrete Bayesian networks , 2015, The Annals of Statistics.

[14]  Bernhard Schölkopf,et al.  Distinguishing Cause from Effect Using Observational Data: Methods and Benchmarks , 2014, J. Mach. Learn. Res..

[15]  Robin J. Evans,et al.  Graphs for Margins of Bayesian Networks , 2014, 1408.1809.

[16]  Peter Bühlmann,et al.  CAM: Causal Additive Models, high-dimensional order search and penalized regression , 2013, ArXiv.

[17]  B. Schölkopf,et al.  Causal discovery with continuous additive noise models , 2013, J. Mach. Learn. Res..

[18]  P. Smyth,et al.  Cyclic Causal Discovery from Continuous Equilibrium Data , 2013, UAI.

[19]  Frederick Eberhardt,et al.  Discovering Cyclic Causal Models with Latent Variables: A General SAT-Based Procedure , 2013, UAI.

[20]  J. Mooij,et al.  From Ordinary Differential Equations to Structural Causal Models: the deterministic case , 2013, UAI.

[21]  M. Drton,et al.  Half-trek criterion for generic identifiability of linear structural equation models , 2011, 1107.5552.

[22]  Frederick Eberhardt,et al.  Combining Experiments to Discover Linear Cyclic Models with Latent Variables , 2010, AISTATS.

[23]  Jiji Zhang,et al.  On the completeness of orientation rules for causal discovery in the presence of latent confounders and selection bias , 2008, Artif. Intell..

[24]  M. Maathuis,et al.  Estimating high-dimensional intervention effects from observational data , 2008, 0810.4214.

[25]  Patrik O. Hoyer,et al.  Discovering Cyclic Causal Models by Independent Components Analysis , 2008, UAI.

[26]  Judea Pearl,et al.  Complete Identification Methods for the Causal Hierarchy , 2008, J. Mach. Learn. Res..

[27]  Kevin P. Murphy,et al.  Exact Bayesian structure learning from uncertain interventions , 2007, AISTATS.

[28]  T. Richardson.  Markov Properties for Acyclic Directed Mixed Graphs , 2003 .

[29]  A. Dawid.  Influence Diagrams for Causal Modelling and Inference , 2002 .

[30]  P. Spirtes,et al.  Ancestral graph Markov models , 2002 .

[31]  Jin Tian,et al.  Causal Discovery from Changes , 2001, UAI.

[32]  Radford M. Neal.  On Deducing Conditional Independence from d-Separation in Causal Graphs with Feedback (Research Note) , 2000, J. Artif. Intell. Res..

[33]  J. Koster.  On the Validity of the Markov Interpretation of Path Diagrams of Gaussian Structural Equations Systems with Correlated Errors , 1999 .

[34]  P. Spirtes,et al.  Using Path Diagrams as a Structural Equation Modeling Tool , 1998 .

[35]  Joseph Y. Halpern.  Axiomatizing Causal Reasoning , 1998, UAI.

[36]  Gregory F. Cooper,et al.  A Simple Constraint-Based Algorithm for Efficiently Mining Observational Databases for Causal Relationships , 1997, Data Mining and Knowledge Discovery.

[37]  J. Koster,et al.  Markov properties of nonrecursive causal models , 1996 .

[38]  Rina Dechter,et al.  Identifying Independencies in Causal Graphs with Feedback , 1996, UAI.

[39]  Thomas S. Richardson,et al.  A Discovery Algorithm for Directed Cyclic Graphs , 1996, UAI.

[40]  Christopher Meek,et al.  Strong completeness and faithfulness in Bayesian networks , 1995, UAI.

[41]  Peter Spirtes,et al.  Directed Cyclic Graphical Representations of Feedback Models , 1995, UAI.

[42]  Thomas S. Richardson,et al.  Causal Inference in the Presence of Latent Variables and Selection Bias , 1995, UAI.

[43]  Judea Pearl,et al.  Probabilistic Evaluation of Counterfactual Queries , 1994, AAAI.

[44]  Herbert A. Simon,et al.  Causality and Model Abstraction , 1994, Artif. Intell..

[45]  Steffen L. Lauritzen,et al.  Independence properties of directed markov fields , 1990, Networks.

[46]  Judea Pearl,et al.  A Constraint-Propagation Approach to Probabilistic Reasoning , 1985, UAI.

[47]  D. A. Kenny,et al.  Correlation and Causation , 1937, Wilmott.

[48]  David Lewis.  Counterfactual Dependence and Time's Arrow , 1979 .

[49]  D. Rubin.  Estimating causal effects of treatments in randomized and nonrandomized studies. , 1974 .

[50]  Arthur S. Goldberger,et al.  Structural Equation Models in the Social Sciences. , 1974 .

[51]  S. J. Mason.  Feedback Theory-Further Properties of Signal Flow Graphs , 1956, Proceedings of the IRE.

[52]  Samuel J. Mason,et al.  Feedback Theory-Some Properties of Signal Flow Graphs , 1953, Proceedings of the IRE.

[53]  Atrick,et al.  SUPPLEMENT TO “FOUNDATIONS OF STRUCTURAL CAUSAL MODELS WITH CYCLES AND LATENT VARIABLES” , 2021 .

[54]  Niels Richard Hansen,et al.  Causal Learning for Partially Observed Stochastic Dynamical Systems , 2018, UAI.

[55]  Andreas Ritter,et al.  Structural Equations With Latent Variables , 2016 .

[56]  T. Richardson.  Single World Intervention Graphs ( SWIGs ) : A Unification of the Counterfactual and Graphical Approaches to Causality , 2013 .

[57]  Frederick Eberhardt,et al.  Learning linear cyclic causal models with latent variables , 2012, J. Mach. Learn. Res..

[58]  Peter Spirtes,et al.  Directed Cyclic Graphs, Conditional Independence, and Non-Recursive Linear Structural Equation Models , 2005 .

[59]  Gregory F. Cooper,et al.  A bayesian local causal discovery framework , 2005 .

[60]  J. Pearl,et al.  Studies in causal reasoning and learning , 2002 .

[61]  J. Pearl.  Causality: Models, Reasoning and Inference , 2000 .

[62]  N. Roese.  Counterfactual thinking. , 1997, Psychological bulletin.

[63]  Thomas S. Richardson,et al.  Automated discovery of linear feedback models , 1996 .

[64]  T. Richardson.  Discovering cyclic causal structure , 1996 .

[65]  Peter Spirtes,et al.  Conditional Independence in Directed Cyclic Graphical Models for Feedback , 1994 .

[66]  A. P. Dawid,et al.  Independence properties of directed Markov fields. Networks, 20, 491-505 , 1990 .

[67]  F. Fisher.  A Correspondence Principle for Simultaneous Equation Models , 1970 .

[68]  T. Haavelmo.  The Statistical Implications of a System of Simultaneous Equations , 1943 .