Causal inference from data. On some inadequacy problems of structures with hidden causes

O.S. Balabanov

Abstract


The reliability of causal inference from data (by independence-based methods) is analyzed. We uncover some mechanisms which may result in model inadequacy due to sample bias and hidden variables. We detect some specific problems in recognition of direction of influence when some causes are hidden. Incorrectness of known rule for edge orientation (under causal insufficiency) is revealed. We suggest the correction to the rule aiming to retain model adequacy.

Problems in programming 2020; 2-3: 392-406


Keywords


causal network; edge orientation; conditional independence; d-separation; collider; chain

Full Text:

PDF (Ukrainian)

References


Pearl J. (2000). Causality: models, reasoning, and inference. – Cambridge: Cambridge Univ. Press. 526 p.

Spirtes P., Glymour C., Scheines R. (2001) Causation, prediction and search. – New York: MIT Press. 543 p.

Balabanov O.S. (2019) Big Data Analytics: principles, trends and tasks (a survey). Problems in Programming. 2019. (3), 47–68. (ISSN 1727–4907) [in Ukrainian]

Spirtes P., Zhang K. (2016) Causal discovery and inference: concepts and recent methodological advances. Applied Informatics. V.3: 3. 28 p.

Spirtes P., Meek C., Richardson T. (1999) An algorithm for causal inference in the presence of latent variables and selection bias. In: Computation, Causation, and Discovery, C. Glymour, and G. Cooper (eds.), P. 211–252. – AAAI Press, Menlo Park, CA. 568 p.

Balabanov O.S. (2016) Induced Dependence, Factor Interaction, and Discriminating between Causal Structures. Cybernetics and Systems Analysis. V. 52. Issue 1. – P. 8–19.

Zhang J. (2008) On the completeness of orientation rules for causal discovery in the presence of latent confounders and selection bias. Artificial Intelligence. V. 172. P. 1873–1896.

Balabanov O.S. (2014). Causal nets: analysis, synthesis and inference from statistical data, Doctor of math. sciences thesis, V.M. Glushkov Institute of Cybernetics, Kyiv, Ukraine. [In Ukrainian]

Balabanov A.S. (2008). Minimal separators in dependency structures: Properties and identification. Cybernetics and Systems Analysis. V. 44, Issue 6, P. 803–815.

O. S. Balabanov. (2013) Logic of minimal separation in causal networks. Cybernetics and Systems Analysis. V. 49. – Issue 2, P. 191–200.

Balabanov O.S. (2017) Principles and analytical tools for reconstruction of probabilistic dependency structures in special class. Problems in programming. 2017. N 1. P. 97–110. [in Ukrainian]

Meek C. (1995) Causal inference and causal explanation with background knowledge. Proceedings of the 11th Conf. on Uncertainty in Artificial Intelligence, P. Besnard and S. Hanks (Eds.), Morgan Kaufmann Publ., Inc., San Mateo, CA. P. 403–410.

Richardson T., Spirtes P. (2002) Ancestral graph Markov models. The Annals of Statistics. V. 30. N 4. P. 962–1030.


Refbacks

  • There are currently no refbacks.