A Bayesian Approach to Causal Discovery

By David E. Heckerman, Christopher Meek, Gregory F. Cooper. 2006

306 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Networks and Communications, Management Science and Operations Research

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Networks and Communications

In Management Science and Operations Research

Citations in the corpus

Causal Discovery from Heterogeneous/Nonstationary Data (2019, background): “, 1995) searches for the equivalence class which gives the highest score under some scoring criteria, such as the Bayesian Information Criterion (BIC), the posterior of the graph given the data (Heckerman et al., 2006), and the generalized score functions (Huang et al.”
Invariant Causal Prediction for Nonlinear Models (2017, methodology): “Heckerman (1997) describe a score-based method with a Bayesian score.”
Causal Inference Using the Algorithmic Markov Condition (2008, neutral): “42”
Telling cause from effect based on high-dimensional observations (2009, methodology): “Before proceeding to describe our method, we should mention connections to Bayesian approaches (Heckerman et al., 1999) to causal discovery.”
A theory of causal learning in children: causal maps and Bayes nets. (2004, methodology): “These learning procedures include Bayesian methods ( Heckerman, Meek, & Cooper, 1999 ) constraint-based methods (Scheines, Spirtes, Glymour, & Meek, 1994), and various combinations of the two.”
Causal Inference on Discrete Data Using Additive Noise Models (2009, background): “[6] are also applicable, but the construction of good priors is 2436 IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, VOL.”
Active Learning of Causal Networks with Intervention Experiments and Optimal Designs (2008, methodology): “There are many methods of structural learning, and the main methods are Bayesian methods (Cooper and Yoo, 1999; Heckerman, 1997) and constraint-based methods (Spirtes et al.”
A Theoretical Study of Y Structures for Causal Discovery (2006, background): “…the large sample limit, but it is a constraint-based search, and such methods have several important disadvantages relative to score-based searches (Heckerman et al., 1999)—(1) the inability to include prior belief in the form of structure probabilities, (2) the output is based on the signiﬁcance…”
Identifying Cause and Effect on Discrete Data using Additive Noise Models (2010, background): “For discrete variables, several Bayesian approaches (Heckerman et al., 1999) are also applicable, but the construction of good priors are challenging and often the latter are designed such that Markov equivalent DAGs still remain indis- tinguishable.”
Detection of Unfaithfulness and Robust Causal Inference (2008, background): “1993; Meek 1995b), to the effect that, from a Bayesian perspective, such an exact cancellation has zero probability under any ‘‘smooth’’(13) probability distribution—and hence any of the priors commonly used in practice (Heckerman et al. 1999)—over the parameters that represent the strength of each ”
Causal Confusion in Imitation Learning (2019, background): “e general problem of deducing cause-effect relationships among variables [50, 36, 38, 48, 9, 49]. “Causal discovery” approaches allow causal inference from pre-recorded observations under constraints [52, 15, 27, 13, 28, 29, 24, 12, 32, 55]. Observational causal inference is known to be impossible i”
From Dependence to Causation (2016, methodology): “Score methods Score methods (Heckerman et al., 1997) construct a mapping from parameter vectors θ ∈ Rm to the set of DAGs on d nodes, and evaluate the score of each candidate by using the posterior distribution p(θ = θ|x1, . . . , xn) = p(x1, . . . , xn|θ = θ)p(θ = θ) p(x1, . . . , xn) ,…”
Adjacency-Faithfulness and Conservative Causal Inference (2006, background): “…causal discovery algorithms assume the causal Markov condition, and most of them (e.g., those discussed in Pearl 2000, Spirtes et al. 2000, Heckerman et al. 1999) also assume, if only implicitly, the converse principle, known as the Causal Faithfulness or Stability Condition: Causal…”
Advances in Exact Bayesian Structure Discovery in Bayesian Networks (2006, background): “In causal discovery, for example, one aims at the identification of (direct) causal relations among a set of attributes, which are qualitatively represented by edges of a Bayesian network (see, e.g., Cooper and Herskovits, 1992, Heckerman et al., 1999, Pearl, 2000, Spirtes et al., 2000).”
Optimal experimental design via Bayesian optimization: active causal structure learning for Gaussian process networks (2019, methodology): “We thus adopt a Bayesian approach to causal discovery [7, 8].”
Causal Network Learning from Multiple Interventions of Unknown Manipulated Targets (2016, methodology): “Many methods have been developed to learn the structures of DAGs from observational and/or experimental data (Cooper and Yoo, 1999; Finegold and Drton, 2011; Friedman, 2004; Geng et al., 2004; Heckerman et al., 1999; Jansen et al., 2003; Maathuis et al., 2009; Pearl, 1995).”
A Survey of Learning Causality with Data: Problems and Methods (2018, background): “Historically, learning causality has been studied in myriad of high-impact domains including education [35, 57, 59, 75], medical science [89, 97], economics [66], epidemiology [58, 87, 115], meteorology [36] and environmental health [81].”