Causal Inference Using Nonnormality

By Yutaka Kano, Shohei Shimizu. 2004

86 citations recorded by the source corpus.

How researchers see this paper

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

Citations in the corpus

Detecting non-causal artifacts in multivariate linear regression models (2018, background): “In addition to the more ‘traditional’ causal Markov condition and causal faithfulness assumption Spirtes et al. (1993); Pearl (2000), researchers have also stated assumptions that admit causal inference when no conditional statistical independences hold, e.”
Algorithmic independence of initial condition and dynamical law in thermodynamics and causal inference (2015, background): “[38] infers the direction with additive independent noise to be the causal one.”
Distinguishing Cause from Effect Using Observational Data: Methods and Benchmarks (2014, methodology): “This means that whenever we can model an observational distribution PX,Y with a model of the form (3), we can also model it using (2), and therefore the causal relationship between X and Y is not identifiable from the observational distribution without making additional assumptions.”
Causal Inference Using the Algorithmic Markov Condition (2008, background): “The authors of [14] observe that there are joint distributions of X1, .”
Telling cause from effect based on high-dimensional observations (2009, background): “However, the fact that deterministic dependencies exist in real-world settings shows that priors that are densities on the parameters of the Bayesian networks, as it is usually assumed, are problematic, and the construction of good priors becomes difficult.”
Testing whether linear equations are causal: A free probability theory approach (2011, background): “In particular one can infer – with some reliability – whether X causes Y or Y causes X, given that either statement is true (e.g. Friedman and Nachman (2000); Kano and Shimizu (2003); Sun et al. (2006); Hoyer et al. (2009); Daniušis et al. (2010); Mooij et al. (2009)).”
Causal Inference on Discrete Data Using Additive Noise Models (2009, methodology): “Kano & Shimizu (2003) use models where f is a linear function and N is additive noise that is independent of the hypothetical cause X .”
Cause-Effect Inference by Comparing Regression Errors (2018, background): “In particular, the identification of the causal relation between only two observed variables is a challenging task [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15].”
Causal Inference by Choosing Graphs with Most Plausible Markov Kernels (2006, background): “The intuition behind such entropy maximization is to complete the linear part of the e(cid:11)ect of Pa j on X j by (cid:12)xing it with the maximal uncertainty, since we consider, like the authors of [4], a linear causal relationship as a simplest form of causality.”
On the Entropy Production of Time Series with Unidirectional Linearity (2009, background): “In the context of causal inference from statistical data, it has been proposed to consider the direction of the linear model as the causal direction [14, 15].”
Joint Causal Inference from Multiple Contexts (2016, background): “, 2014), algebraic constraints in the linear-Gaussian case (van Ommen and Mooij, 2017), non-Gaussianity in linear models (Kano and Shimizu, 2003), and non-additivity of noise in nonlinear models (Peters et al.”
Causal reasoning by evaluating the complexity of conditional densities with kernel methods (2008, background): “A related idea for causal inference was described by Kano and Shimizu in [8].”
Distinguishing Cause from Effect Using Observational Data: Methods and Benchmarks (2014, background): “The work of Kano and Shimizu (2003); Shimizu et al. (2006) showed that for linear models (i.e., where the functions fX and fY are restricted to be linear), non-Gaussianity of the input and noise distributions actually allows one to distinguish the directionality of such functional models.”
Identifying Cause and Effect on Discrete Data using Additive Noise Models (2010, unknown): “Kano & Shimizu (2003) and Shimizu et al. (2006) use models Y = f (X) + N (1) where f is a linear function and N is additive noise that is independent of the hypothetical cause X.”
Distinguishing Cause and Effect via Second Order Exponential Models (2009, background): “[5] have observed that linear causal relationships between non-Gaussian distributed random variables induce joint measures which require non-linear cause-eﬀect relations for the wrong causal directions.”
Distinguishing between cause and effect (2008, background): “Recent work (Kano and Shimizu, 2003; Sun et al., 2006; Shimizu et al., 2006; Sun et al., 2008; Hoyer et al., 2009; Janzing and Schölkopf, 2008) suggests that the shape of the joint distribution shows asymmetries between cause and effect, which often indicates the causal direction with some…”
UvA-DARE ( Digital Academic Repository ) Causal Discovery with Continuous Additive Noise Models (2014, unknown): “3 Estimating the DAG: LiNGAM Kano and Shimizu (2003) and Shimizu et al.”