Abstract
Detecting causal relationships between random variables using only matched pairs of noisy observations is a crucial problem in many scientific fields. In this paper the problem is addressed by extracting a number of features for each matched pair using a selection of curve-fitting and information theoretic features. Using these features, we train a pair of Gradient Boosting Machines whose hyperparameters we optimise using stochastic simultaneous optimistic optimisation. The results show that our method is relatively successful, gaining a third place in the 2013 Kaggle’s Causality Challenge. Our method is sound enough to be used in causality detection (or as part of a more comprehensive toolkit), although we believe it might be possible to considerably improve the quality of results by adding more features in the same vein.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
David Arthur and Sergei Vassilvitskii. k-means++: The advantages of careful seeding. In Proceedings of the eighteenth annual ACM-SIAM symposium on Discrete algorithms, pages 1027–1035. Society for Industrial and Applied Mathematics, 2007.
Andrea Falcon. Aristotle on causality. In Edward N. Zalta, editor, The Stanford Encyclopedia of Philosophy. Winter 2012 edition, 2012.
James George Frazer. The Golden Bough: A Study in Magic and Religion. Vol. 13, Aftermath: a Supplement to the Golden Bough. Macmillan, 1936.
Jerome H Friedman. Greedy function approximation: a gradient boosting machine. Annals of Statistics, pages 1189–1232, 2001.
Patrik O Hoyer, Dominik Janzing, Joris M Mooij, Jonas Peters, and Bernhard Schölkopf. Nonlinear causal discovery with additive noise models. 2009.
Dominik Janzing, Joris Mooij, Kun Zhang, Jan Lemeire, Jakob Zscheischler, Povilas Daniušis, Bastian Steudel, and Bernhard Schölkopf. Information-geometric approach to inferring causal directions. Artificial Intelligence, 182:1–31, 2012.
Judea Pearl. Causality: models, reasoning and inference, volume 29. Cambridge Univ Press, 2000.
Shohei Shimizu, Patrik O Hoyer, Aapo Hyvärinen, and Antti Kerminen. A linear non-Gaussian acyclic model for causal discovery. The Journal of Machine Learning Research, 7:2003–2030, 2006.
Michal Valko, Alexandra Carpentier, and Rémi Munos. Stochastic Simultaneous Optimistic Optimization. In 30th International Conference on Machine Learning, Atlanta, États-Unis, February 2013. URL http://hal.inria.fr/hal-00789606.
Acknowledgment
This work was supported by EPSRC grant EP/H048588/1 entitled: “UCT for Games and Beyond”.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Appendix: Causality Challenge
Appendix: Causality Challenge
Title: Training Gradient Boosting Machines using Curve-fitting and Information theoretic features for Causal Direction Detection.
Participant name, address, email and website: Spyridon Samothrakis, Diego Perez, https://github.com/ssamot/causality.
Task(s) solved: Kaggle Competition.
Reference: This paper.
Method: A combination of feature extraction from the sample data, Gradient boosting machines and StoSOO meta-optimisation.
-
Preprocessing: Exploit Symmetries.
-
Causal discovery: Gradient Boosting Machine, Curve fitting/Information theoretic features.
-
Feature selection: Feature Ranking.
-
Classification: Gradient Boosting Machine
-
Model selection/hyperparameter selection: Cross-validation, Stochastic Simultaneous Optimistic Optimisation.
Results (Table 11.1 ):
-
quantitative advantages: The method and ideas behind our method are relatively simple. We advocate a feature extraction strategy based on curve fitting + information theoretic features.
-
qualitative advantages: There are some elements of novelty, mostly in the ideas behind extracting features and doing hyper-parameter optimisation.
Code and installation instructions can be found here: https://github.com/ssamot/causality.
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this chapter
Cite this chapter
Samothrakis, S., Perez, D., Lucas, S. (2019). Training Gradient Boosting Machines Using Curve-Fitting and Information-Theoretic Features for Causal Direction Detection. In: Guyon, I., Statnikov, A., Batu, B. (eds) Cause Effect Pairs in Machine Learning. The Springer Series on Challenges in Machine Learning. Springer, Cham. https://doi.org/10.1007/978-3-030-21810-2_11
Download citation
DOI: https://doi.org/10.1007/978-3-030-21810-2_11
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-21809-6
Online ISBN: 978-3-030-21810-2
eBook Packages: Computer ScienceComputer Science (R0)