Do we become wiser with time? On causal equivalence with tiered background knowledge

Equivalence classes of DAGs (represented by CPDAGs) may be too large to provide useful causal information. Here, we address incorporating tiered background knowledge yielding restricted equivalence classes represented by 'tiered MPDAGs'. Tiered knowledge leads to considerable gains in informativeness and computational efficiency: We show that construction of tiered MPDAGs only requires application of Meek's 1st rule, and that tiered MPDAGs (unlike general MPDAGs) are chain graphs with chordal components. This entails simplifications e.g. of determining valid adjustment sets for causal effect estimation. Further, we characterise when one tiered ordering is more informative than another, providing insights into useful aspects of background knowledge.

[1]  Tian-Zuo Wang,et al.  Sound and Complete Causal Identification with Latent Variables Given Local Background Knowledge , 2023, NeurIPS.

[2]  Yangbo He,et al.  On the Representation of Causal Background Knowledge and its Applications in Causal Inference , 2022, arXiv.org.

[3]  V. Didelez,et al.  A longitudinal causal graph analysis investigating modifiable risk factors and obesity in a European cohort of children and adolescents , 2022, medRxiv.

[4]  M. Liskiewicz,et al.  Polynomial-Time Algorithms for Counting and Sampling Markov Equivalent DAGs with Applications , 2022, J. Mach. Learn. Res..

[5]  M. Osler,et al.  Data-Driven Model Building for Life Course Epidemiology. , 2021, American journal of epidemiology.

[6]  Richard Guo,et al.  Minimal enumeration of all possible total effects in a Markov equivalence class , 2020, AISTATS.

[7]  M. Maathuis,et al.  On efficient adjustment in causal graphs , 2020, 2002.06825.

[8]  M. Maathuis,et al.  Graphical criteria for efficient total effect estimation via adjustment in causal linear models , 2019, 1907.02435.

[9]  Bernhard Schölkopf,et al.  Inferring causation from time series in Earth system sciences , 2019, Nature Communications.

[10]  Marloes H. Maathuis,et al.  Interpreting and Using CPDAGs With Background Knowledge , 2017, UAI.

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

[12]  Johannes Textor,et al.  Complete Graphical Characterization and Construction of Adjustment Sets in Markov Equivalence Classes of Ancestral Graphs , 2016, J. Mach. Learn. Res..

[13]  Maciej Liskiewicz,et al.  Separators and Adjustment Sets in Markov Equivalent DAGs , 2016, AAAI.

[14]  M. Maathuis,et al.  Estimating the effect of joint interventions from observational data in sparse high-dimensional settings , 2014, 1407.2451.

[15]  Peter Bühlmann,et al.  Two optimal strategies for active learning of causal models from interventional data , 2012, Int. J. Approx. Reason..

[16]  Peter Bühlmann,et al.  Characterization and Greedy Learning of Interventional Markov Equivalence Classes of Directed Acyclic Graphs (Abstract) , 2011, UAI.

[17]  P. Spirtes,et al.  MARKOV EQUIVALENCE FOR ANCESTRAL GRAPHS , 2009, 0908.3605.

[18]  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..

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

[20]  Frederick Eberhardt,et al.  Almost Optimal Intervention Sets for Causal Discovery , 2008, UAI.

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

[22]  John Fox,et al.  A Life Course Approach to Chronic Disease Epidemiology , 1998, BMJ.

[23]  D. Madigan,et al.  A characterization of Markov equivalence classes for acyclic digraphs , 1997 .

[24]  Christopher Meek,et al.  Causal inference and causal explanation with background knowledge , 1995, UAI.

[25]  Judea Pearl,et al.  Equivalence and Synthesis of Causal Models , 1990, UAI.

[26]  Illtyd Trethowan.  Causality , 1938 .

[27]  Zhuangyan Fang,et al.  IDA with Background Knowledge , 2020, UAI.

[28]  Bryan Andrews,et al.  On the Completeness of Causal Discovery in the Presence of Latent Confounding with Tiered Background Knowledge , 2020, AISTATS.

[29]  Daniel Malinsky,et al.  Causal Structure Learning from Time Series Causal Structure Learning from Multivariate Time Series in Settings with Unmeasured Confounding , 2018 .

[30]  R Scheines,et al.  The TETRAD Project: Constraint Based Aids to Causal Model Specification. , 1998, Multivariate behavioral research.