Group invariance principles for causal generative models

By Michel Besserve, Naji Shajarisales, Bernhard Schölkopf, Dominik Janzing. 2017

50 citations recorded by the source corpus.

The postulate of independence of cause and mechanism (ICM) has recently led to several new causal discovery algorithms. The interpretation of independence and the way it is utilized, however, varies across these methods. Our aim in this paper is to propose a group theoretic framework for ICM to unify and generalize these approaches. In our setting, the cause-mechanism relationship is assessed by comparing it against a null hypothesis through the application of random generic group transformations. We show that the group theoretic view provides a very general tool to study the structure of data generating mechanisms with direct applications to machine learning.

Subfields: Artificial Intelligence, Molecular Biology, Generative Models, Causal Inference & Discovery

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 Molecular Biology

In Generative Models

In Causal Inference & Discovery

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, background): “Genericity was originally deﬁned by Besserve et al. (2018) as a measure of independence between cause V = f 1 ( Z ) and mechanism f 2 .”
Causality for Machine Learning (2019, background): “…or mechanisms, we discuss the special case of intervening upon the latent noise variables.19 One way to intervene is to replace noise variables with the corresponding values computed from other input images, a procedure that has been referred to as hybridization by Besserve et al. (2018b).”
Counterfactuals uncover the modular structure of deep generative models (2018, background): “This principle can be applied to generative models to assess how well they capture a causal mechanism (Besserve et al., 2018).”
Generalization in anti-causal learning (2018, background): “Experiments: Besserve et al. (2017, 2018) study properties of learned causal generative models that distinguish them from anti-causal models such as encoders or discriminators.”
Conditional variance penalties and domain shift robustness (2017, background): “the problems of causeeffect inference and generative adversarial networks (Chalupka et al., 2014; Lopez-Paz et al., 2017; Lopez-Paz and Oquab, 2017; Goudet et al., 2017; Bahadori et al., 2017; Besserve et al., 2018; Kocaoglu et al., 2018).”
CausalGAN: Learning Causal Implicit Generative Models with Adversarial Training (2017, background): “In a very recent work [5], authors point out the connection of GANs to causal generative models.”
Intrinsic disentanglement: an invariance view for deep generative models (2018, methodology): “This approach relates to the concept of ICM, in the sense that the family of transformations T ′ decouples the cause V3 from the mechanism m (Besserve et al., 2018).”
Guarding Against Adversarial Domain Shifts with Counterfactual Regularization (2017, background): “The latter estimates are obtained by causality detection networks or scores such as estimated by the NCC. Besserve et al. (2017) draw connections between GANs and causal generative models, using a group theoretic framework.”