ABSTRACT
By taking into account the nonlinear effect of the cause, the inner noise effect, and the measurement distortion effect in the observed variables, the post-nonlinear (PNL) causal model has demonstrated its excellent performance in distinguishing the cause from effect. However, its identifiability has not been properly addressed, and how to apply it in the case of more than two variables is also a problem. In this paper, we conduct a systematic investigation on its identifiability in the two-variable case. We show that this model is identifiable in most cases; by enumerating all possible situations in which the model is not identifiable, we provide sufficient conditions for its identifiability. Simulations are given to support the theoretical results. Moreover, in the case of more than two variables, we show that the whole causal structure can be found by applying the PNL causal model to each structure in the Markov equivalent class and testing if the disturbance is independent of the direct causes for each variable. In this way the exhaustive search over all possible causal structures is avoided.
- Achard, S., & Jutten, C. (2005). Identifiability of post-nonlinear mixtures. IEEE Signal Processing Letters, 12, 423--426.Google ScholarCross Ref
- Braun, M. (1993). Differential equations and their applications: An introduction to applied mathematics. New York: Springer-Verlag. 4th edition.Google Scholar
- Friedman, N., & Nachman, I. (2000). Gaussian process networks. Proc. 16th Conference on Uncertainty in Artificial Intelligence (UAI) (pp. 211--219). Morgan Kaufmann. Google ScholarDigital Library
- Gretton, A., Fukumizu, K., Teo, C., Song, L., Schölkopf, B., & Smola, A. (2008). A kernel statistical test of independence. NIPS 20 (pp. 585--592). Cambridge, MA: MIT Press.Google Scholar
- Hoyer, P., Janzing, D., Mooji, J., Peters, J., & Schöölkopf, B. (2009). Nonlinear causal discovery with additive noise models. NIPS 21. Vancouver, B.C., Canada.Google Scholar
- Hyvärinen, A., Karhunen, J., & Oja, E. (2001). Independent component analysis. John Wiley & Sons, Inc.Google Scholar
- Lin, J. (1998). Factorizing multivariate function classes. NIPS 10 (pp. 563--569). Cambridge, MA: MIT Press. Google ScholarDigital Library
- Mooij, J., Janzing, D., & Schöölkopf, B. (2008). Distinguishing between cause and effect. URL http://www.kyb.tuebingen.mpg.de/bs/people/jorism/causality-data/.Google Scholar
- Pearl, J. (2000). Causality: Models, reasoning, and inference. Cambridge: Cambridge University Press. Google ScholarDigital Library
- Polyanin, A., & Zaitsev, V. (2003). Handbook of exact solutions for ordinary differential equations. Boca Raton: Chapman & Hall/CRC Press. 2nd edition.Google Scholar
- Polyanin, A., & Zaitsev, V. (2004). Handbook of nonlinear partial differential equations (supplement s.4.4). Boca Raton: Chapman & Hall/CRC Press.Google Scholar
- Shimizu, S., Hoyer, P., Hyvärinen, A., & Kerminen, A. (2006). A linear non-Gaussian acyclic model for causal discovery. Journal of Machine Learning Research, 7, 2003--2030. Google ScholarDigital Library
- Spirtes, P., Glymour, C., & Scheines, R. (2001). Causation, prediction, and search. Cambridge, MA: MIT Press. 2nd edition.Google Scholar
- Sun, X., Janzing, D., & Schöölkopf, B. (2008). Causal reasoning by evaluating the complexity of conditional densities with kernel methods. Neurocomputing, 71, 1248--1256. Google ScholarDigital Library
- Verma, T., & Pearl, J. (1990). Equivalence and synthesis of causal models. Proceedings of 6th Conference on Uncertainty in Artificial Intelligence (UAI) (pp. 220--227). Boston, MA: Morgan Kaufmann. Google ScholarDigital Library
- Zhang, K., & Hyväärinen, A. (2008). Distinguishing causes from effects using nonlinear acyclic causal models. Presented at NIPS 2008 Workshop on Causality. URL http://www.cs.helsinki.fi/u/ahyvarin/papers/Zhang09NIPSworkshop.pdf.Google Scholar
Index Terms
- On the identifiability of the post-nonlinear causal model
Recommendations
Causal effect identifiability under partial-observability
ICML'20: Proceedings of the 37th International Conference on Machine LearningCausal effect identifiability is concerned with establishing the effect of intervening on a set of variables on another set of variables from observational or interventional distributions under causal assumptions that are usually encoded in the form of a ...
Identifiability in causal Bayesian networks: a sound and complete algorithm
AAAI'06: proceedings of the 21st national conference on Artificial intelligence - Volume 2This paper addresses the problem of identifying causal effects from nonexperimental data in a causal Bayesian network, i.e., a directed acyclic graph that represents causal relationships. The identifiability question asks whether it is possible to ...
Identifiability of Causal Effects in a Multi-Agent Causal Model
IAT '03: Proceedings of the IEEE/WIC International Conference on Intelligent Agent TechnologyThis paper is a first step to extending Judea Pearl's work on identification of causal effects to a multi-agent context. We introduce multi-agent causal models consisting of a collection of agents each having access to a non-disjoint subset of the ...
Comments