Conditional distribution variability measures for causality detection

By José A. R. Fonollosa. 2016

51 citations recorded by the source corpus.

In this paper we derive variability measures for the conditional probability distributions of a pair of random variables, and we study its application in the inference of causal-effect relationships. We also study the combination of the proposed measures with standard statistical measures in the the framework of the ChaLearn cause-effect pair challenge. The developed model obtains an AUC score of 0.82 on the final test database and ranked second in the challenge.

Subfields: Artificial Intelligence, Statistics and Probability

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

1. Yee Whye Teh
2. Jovana Mitrovi'c
3. Dino Sejdinovic
4. Diviyan Kalainathan
5. David Lopez-Paz
6. Isabelle M Guyon
7. Michèle Sebag
8. Olivier Goudet

In Statistics and Probability

1. David Lopez-Paz
2. Dino Sejdinovic
3. Yee Whye Teh
4. Isabelle M Guyon
5. Olivier Goudet
6. Jovana Mitrovi'c
7. Michèle Sebag
8. Diviyan Kalainathan

Citations in the corpus

• Causal Generative Neural Networks (2017, background): “…rely on different but equally strong assumptions, such as linearity [Shimizu et al. , 2006], additive noise [Zhang and Hyv¨arinen, 2009; Peters et al. , 2014], determinism [Daniusis et al. , 2012], or a large corpus of annotated causal relations [Lopez-Paz et al. , 2015; Fonollosa, 2016].”
• Explainable and Interpretable Models in Computer Vision and Machine Learning (2017, unknown): “…selected as causal direction; vi) CDS, retaining the causal orientation with lowest variance of the conditional probability distribution; vii) Jarfo (Fonollosa, 2016), using a random forest causal classifier trained from the ChaLearn Cause-effect pairs on top of 150 features including ANM, IGCI,…”
• Causal Discovery Toolbox: Uncover causal relationships in Python (2019, methodology): “…by samples of their joint distribution and label (cid:96) i indicates the type of causal relationship among both variables (independent, The CEP challenges spurred the development of state-of-the-art pairwise causal identification methods such as Jarfo, RCC (Fonollosa, 2016; Lopez-Paz et al., 2015).”
• Causal Inference via Kernel Deviance Measures (2018, methodology): “In [5], the feature representation includes quantities describing the joint, marginal and conditional distributions.”
• Automated Machine Learning: Methods, Systems, Challenges (2019, unknown): “The result is called “Jack-knife correlation coefficient” [13].”