Probabilistic latent variable models for distinguishing between cause and effect

By Joris M. Mooij, Oliver Stegle, Dominik Janzing, Kun Zhang, Bernhard Schölkopf. 2010

142 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

Causal Generative Neural Networks (2017, background): “…exploit distributional asymmetries to discover bivariate causal relations [Hoyer et al. , 2009; Zhang and Hyv¨arinen, 2009; Daniusis et al. , 2012; Stegle et al. , 2010; Lopez-Paz et al. , 2015; Fonollosa, 2016], while others rely on conditional independence to discover structures on three or…”
Telling cause from effect in deterministic linear dynamical systems (2015, methodology): “The idea behind SIC, as well as several new approaches to causal inference (Daniusis et al., 2010; Janzing et al., 2010; Mooij et al., 2010; Zscheischler et al., 2011; Janzing et al., 2012; Sgouritsa et al., 2015), is to rely on the “philosophical” principle that the cause and”
Distinguishing Cause from Effect Using Observational Data: Methods and Benchmarks (2014, positive): “The former conclusion about the performance of ANM-pHSIC is in line with earlier reports, but the latter conclusion is surprising, considering that good performance of IGCI-slope and IGCI-ent-1sp has been reported on several occasions in earlier work (Daniušis et al., 2010; Mooij et al., 2010; Janzi”
SAM: Structural Agnostic Model, Causal Discovery and Penalized Adversarial Learning (2018, methodology): “Examples are the Additive Noise Model or ANM (Hoyer et al., 2009); the Gaussian Process Inference causal model, or GPI (Stegle et al., 2010); or the Randomized Causation Coefﬁ-cient, (RCC) (Lopez-Paz et al., 2015).”
Explainable and Interpretable Models in Computer Vision and Machine Learning (2017, unknown): “( Hoyer et al., 2009; Stegle et al., 2010; Daniusis et al., 2012; Mooij et al., 2016) In this paper, the learning of functional causal models is tackled in the search space of generative neural networks (Kingma and Welling, 2013; Goodfellow et al., 2014), and aims at the functional causal model…”
Single-Model Uncertainties for Deep Learning (2018, background): “ause-effect literature, including LINGAM (Shimizu et al., 2006), GR-AN (HernandezLobato et al., 2016), ANM (Hoyer et al., 2009), PNL (Zhang & Hyv¨arinen, 2009), IGCI (Daniusis et al., 2012), GPI (Stegle et al., 2010), and EMD (Chen et al., 2014). The same table shows the importance of aggregatin”
Distinguishing Cause from Effect Based on Exogeneity (2015, methodology): “Inspired by the settings in [8, 15], we generated the simulated data with the model Y = ( X + bX 3 ) e αE + (1 − α ) E , where X and E were obtained by passing i.i.d. Gaussian samples through power nonlinearities with exponent q , while keeping the original signs.”
Towards a Learning Theory of Cause-Effect Inference (2015, methodology): “This model was later extended into nonlinear additive noise models (Hoyer et al., 2009; Zhang & Hyvärinen, 2009; Stegle et al., 2010; Kpotufe et al., 2013; Peters et al., 2014), which prefer the causal direction under which the alleged cause is independent from the additive residuals of some…”
Pairwise Likelihood Ratios for Estimation of Non-Gaussian Structural Equation Models (2013, methodology): “The methods were compared with three previously published methods: • LiNGAM estimated using ICA, as proposed in the original paper introducing LiNGAM by Shimizu et al. (2006).4 • DirectLiNGAM, specifically the kernel-based version proposed by Shimizu et al.”
Inference of Cause and Effect with Unsupervised Inverse Regression (2015, methodology): “While Postulate 1 is abstract, the aforementioned approaches provide formal-izations by specifying what is meant by independence or information : Janzing and Sch¨olkopf [2010] postulate algorithmic independence of P ( Y | X ) and P ( X ), i.e. zero algorithmic mutual information: I ( P ( X ) : P (…”
Telling Cause from Effect Using MDL-Based Local and Global Regression (2017, background): “[16] derived an inference rule stating that if X causes Y ,”
The Randomized Causation Coefficient (2014, background): “This relates in spirit to the Bayesian causal inference strategy proposed by Stegle et al. (2010), but avoids the need of expensive, approximate inference at test time.”
Structural Agnostic Modeling: Adversarial Learning of Causal Graphs (2018, methodology): “joint distribution, e.g. based on the assumption that p(x)p(yjx) is simpler than p(y)p(xjy) (for some appropriate notion of simplicity) when Xcauses Y (Hoyer et al., 2009; Zhang and Hyv¨arinen, 2010; Mooij et al., 2010). Another stream of work closely related to causal discovery is the causal featur”
Causal Inference via Kernel Deviance Measures (2018, background): “Analogously, for a conditional distribution p ( Y | X ) , the Kolmogorov complexity is Assuming X → Y , the asymmetry notion speciﬁed in terms of factorization complexity can be expressed as which holds up to an additive constant [Stegle et al., 2010].”