ParceLiNGAM: A Causal Ordering Method Robust Against Latent Confounders

Tashiro, Tatsuya; Shimizu, Shohei; Hyvärinen, Aapo; Washio, Takashi

doi:10.1162/neco_a_00533

Cited by 43 publications

(39 citation statements)

References 17 publications

(25 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Algorithms which learn latent variable LiNGAM models (Hoyer et al, 2008; Kawahara et al, 2010; Entner and Hoyer, 2010; Tashiro et al, 2014) allow for the possibility of unmeasured variables. These algorithms exploit assumptions about the causal structure (assumed to be structural equation models which are acyclic, linear, and which have non-Gaussian error terms) to estimate graphical structure and some estimate causal strength parameters simultaneously.…”

Section: Introductionmentioning

confidence: 99%

Estimating bounds on causal effects in high-dimensional and possibly confounded systems

Malinsky

Spirtes

2017

International Journal of Approximate Reasoning

View full text Add to dashboard Cite

We present an algorithm for estimating bounds on causal effects from observational data which combines graphical model search with simple linear regression. We assume that the underlying system can be represented by a linear structural equation model with no feedback, and we allow for the possibility of latent confounders. Under assumptions standard in the causal search literature, we use conditional independence constraints to search for an equivalence class of ancestral graphs. Then, for each model in the equivalence class, we perform the appropriate regression (using causal structure information to determine which covariates to adjust for) to estimate a set of possible causal effects. Our approach is based on the IDA procedure of Maathuis et al. (2009), which assumes that all relevant variables have been measured (i.e., no latent confounders). We generalize their work by relaxing this assumption, which is often violated in applied contexts. We validate the performance of our algorithm in simulation experiments.

show abstract

Section: Introductionmentioning

confidence: 99%

Estimating bounds on causal effects in high-dimensional and possibly confounded systems

Malinsky

Spirtes

2017

International Journal of Approximate Reasoning

View full text Add to dashboard Cite

show abstract

“…In Tashiro et al (2013), DirectLiNGAM was extended in order to be robust against latent confounders. Here, the key concept is to detect latent confounders by testing the independence between estimated exogenous variables, and finding subsets that include variables that are not affected by latent confounders, in order to estimate causal orders one by one, as in DirectLiNGAM.…”

Section: Improvements On the Basic Estimation Methodsmentioning

confidence: 99%

“…For example, non-Gaussianity of exogenous variables can be tested by means of Gaussianity tests for estimated exogenous variables, such as the Kolmogorov-Smirnov test. In addition, violations of the independence of exogenous variables may be detected by using the independence test of residuals (Entner and Hoyer, 2011;Tashiro et al, 2013). The overall suitability of the model assumptions can be evaluated by means of a chi-square test, using higher-order moments (Shimizu and Kano, 2008), although large sample sizes are required in order to estimate higher-order moments accurately.…”

Section: Detection Of Violations Of Model Assumptionsmentioning

confidence: 99%

Lingam: Non-Gaussian Methods for Estimating Causal Structures

Shimizu

2014

Behaviormetrika

Self Cite

View full text Add to dashboard Cite

In many empirical sciences, the causal mechanisms underlying various phenomena need to be studied. Structural equation modeling is a general framework used for multivariate analysis, and provides a powerful method for studying causal mechanisms. However, in many cases, classical structural equation modeling is not capable of estimating the causal directions of variables. This is because it explicitly or implicitly assumes Gaussianity of data and typically utilizes only the covariance structure of data. In many applications, however, non-Gaussian data are often obtained, which means that more information may be contained in the data distribution than the covariance matrix is capable of containing. Thus, many new methods have recently been proposed for utilizing the non-Gaussian structure of data and estimating the causal directions of variables. In this paper, we provide an overview of such recent developments in causal inference, and focus in particular on the non-Gaussian methods known as LiNGAM.

show abstract

“…(The worst-case requires exponential order computation of p, which seems to be very rare.) ParceLiNGAM [19] and LvLiNGAM [5] are major existing approaches to LiNGAM when confounding is present. For the details on them, see the references.…”

Section: Independence Testing and Mutual Informationmentioning

confidence: 99%

“…Some authors have proposed ways to avoid the effects of confounders without extending LiNGAM. However, these methods require the knowledge that confounding is present a priori and take an exponential time of the number p of variables [1,19]. Besides, Shimizu et al [15] considered individual-specific effects that are sometimes the source of confounding, and proposed an empirical Bayesian approach for estimating possible causal direction.…”

Section: Introductionmentioning

confidence: 99%

Causal Order Identification to Address Confounding: Binary Variables

Suzuki

Inaoka

2021

Preprint

View full text Add to dashboard Cite

This paper considers an extension of the linear non-Gaussian acyclic model (LiNGAM) that determines the causal order among variables from a dataset when the variables are expressed by a set of linear equations, including noise. In particular, we assume that the variables are binary. The existing LiNGAM assumes that no confounding is present, which is restrictive in practice. Based on the concept of independent component analysis (ICA), this paper proposes an extended framework in which the mutual information among the noises is minimized. Another significant contribution is to reduce the realization to the shortest path problem, in which the distance between each pair of nodes expresses an associated mutual information value, and the path with the minimum sum (KL divergence) is sought. Although p! mutual information values should be compared, this paper dramatically reduces the computation when no confounding is present. The proposed algorithm finds the globally optimal solution, while the existing approaches locally greedily seek the order based on hypothesis testing. We use the best estimator in the sense of Bayes/MDL that correctly detects independence for mutual information estimation. Experiments using artificial and actual data show that the proposed version of LiNGAM achieves significantly better performance, particularly when confounding is present.

show abstract

ParceLiNGAM: A Causal Ordering Method Robust Against Latent Confounders

Cited by 43 publications

References 17 publications

Estimating bounds on causal effects in high-dimensional and possibly confounded systems

Estimating bounds on causal effects in high-dimensional and possibly confounded systems

Lingam: Non-Gaussian Methods for Estimating Causal Structures

Causal Order Identification to Address Confounding: Binary Variables

Contact Info

Product

Resources

About