Information-geometric approach to inferring causal directions

By Dominik Janzing, Joris M. Mooij, Kun Zhang, Jan Lemeire, Jakob Zscheischler, Povilas Daniusis, Bastian Steudel, Bernhard Schölkopf. 2012

317 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computational Theory and Mathematics, Causal Inference & Discovery, Information Theory

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 Computational Theory and Mathematics

In Causal Inference & Discovery

In Information Theory

Citations in the corpus

Toward Causal Representation Learning (2021, background): “10Other dependence measures have been proposed for highdimensional linear settings and time series [28], [119], [121], [122], [126], [226].”
Towards Causal Representation Learning (2021, background): “9Other dependence measures have been proposed for high-dimensional linear settings and time series [124, 226, 27, 122, 119, 125].”
A theory of independent mechanisms for extrapolation in generative models (2020, methodology): “This comparison is supported by several works on identifiation of causal pairs using IM, showing identifiability can be obtained on toy examples by choosing the direction of causation that maximized genericity (Shajarisales et al. 2015; Zscheischler, Janzing, and Zhang 2011; Janzing, Hoyer, and Schö”
Causality for Machine Learning (2019, background): “…a sum of irregularity already present in the input distribution and irregularity introduced by the function, i.e., the irregularities of the two mechanisms add up rather than compensating each other, which would not be the case in the anti-causal direction (for details, see Janzing et al. (2012)).”
Semi-supervised learning, causality, and the conditional cluster assumption (2019, methodology): “(a) IGCI model for X → Y , taken from [20] XC Y”
Invariant Causal Prediction for Sequential Data (2017, methodology): “Many procedures in this field only use observational data (e.g., Spirtes, Glymour, and Scheines 2000; Chickering 2002; Shimizu et al. 2006; Janzing et al. 2012; van de Geer and Bühlmann 2013; Peters et al. 2014; Bühlmann, Peters, and Ernest 2014).”
Group invariance principles for causal generative models (2017, background): “(8) which corresponds exactly to the orthogonality postulate exploited in Information Geometric Causal Inference (IGCI) (Daniusis et al., 2010; Janzing et al., 2012).”
Algorithmic independence of initial condition and dynamical law in thermodynamics and causal inference (2015, methodology): “As already observed in [49, 50], (4) is equivalent to uncorrelatedness between ¢ f log and pX.”
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., 2012; 2010; Zscheischler et al., 2011), is to rely on the “philosophical” principle that the cause and the mechanism that generates the cause from the effect”
Causal inference by using invariant prediction: identification and confidence intervals (2015, background): “…2012], estimate such a class based on observational or interventional data [Spirtes et al., 2000, Chickering, 2002, Castelo and Kocka, 2003, Kalisch and B¨uhlmann, 2007, He and Geng., 2008, Hauser and B¨uhlmann, 2015, cf.], and ﬁnally infer the iden-tiﬁable causal eﬀects or provide some…”
Distinguishing Cause from Effect Using Observational Data: Methods and Benchmarks (2014, methodology): “There should be better ways of dealing with discretization effects, but we nevertheless include this particular estimator for comparison, as it was also used in previous implementations of the entropy-based IGCI method (Daniušis et al., 2010; Janzing et al., 2012).”
Causal Inference by Stochastic Complexity (2017, methodology): “Altogether we can compute the multinomial stochastic complexity in O(n).”
D’ya Like DAGs? A Survey on Structure Learning and Causal Discovery (2021, background): “The corollary is that there exists dependence between PY and f −1 [112].”
Learning Causal Structures Using Regression Invariance (2017, result): “Similar assumptions have been considered in other work [3, 30, 14, 13, 29, 23].”
Inference of Cause and Effect with Unsupervised Inverse Regression (2015, methodology): “In this case, previous work [Janzing et al., 2012] formalizes “independence” as uncorrelatedness between logf ′ and the density of P (X), both viewed as random variables.”
Invariant Risk Minimization Games (2020, background): “been intervened on, the conditional probability P(yjx Pa(y)) remains invariant. This is called the modularity condition [6]. Related and similar notions are the independent causal mechanism principle [7, 8, 9] and the invariant causal prediction principle [10, 11]. These principles imply that if all”
Robust Tree-based Causal Inference for Complex Ad Effectiveness Analysis (2015, methodology): “Besides propensity-based causal inference methodologies, structural equation model [5], inference based on decomposition of the joint distribution of the treatment and response variables [22, 11], and before-and-after studies [39] are also used for causal inference from different points of view.”
Semi-supervised interpolation in an anticausal learning scenario (2015, methodology): “Our restriction to monotonically increasing bijections of [0] [1] coincides with the typical toy scenario used by Daniusis et al. (2010); Janzing et al. (2012) to explain Information- Geometric Causal Inference (IGCI) although the formalism of IGCI introduced therein is actually more general.”