Abstract
We describe various sets of conditional independence relationships, sufficient for qualitatively comparing non-vanishing squared partial correlations of a Gaussian random vector. These sufficient conditions are satisfied by several graphical Markov models. Rules for comparing degree of association among the vertices of such Gaussian graphical models are also developed. We apply these rules to compare conditional dependencies on Gaussian trees. In particular for trees, we show that such dependence can be completely characterised by the length of the paths joining the dependent vertices to each other and to the vertices conditioned on. We also apply our results to postulate rules for model selection for polytree models. Our rules apply to mutual information of Gaussian random vectors as well.
Similar content being viewed by others
Notes
The author would like to thank the referee for drawing his attention to this equality which improved the original proof immensely.
References
Bekker, P. A., de Leeuw, J. (1987). The rank of reduced dispersion matrices. Psychometrika, 52, 125–135.
Chaudhuri, S. (2005). Using the structure of d-connecting paths as a qualitative measure of the strength of dependence. PhD thesis, Seattle: Department of Statistics, University of Washington.
Chaudhuri, S. (2013). Qualitative inequalities for squared partial correlations of a gaussian random vector. Technical Report 1/2013, Department of Statistics and Applied Probability, National University of Singapore.
Chaudhuri, S., Richardson, T. S. (2003). Using the structure of d-connecting paths as a qualitative measure of the strength of dependence. In: Proceedings of the Nineteenth Conference Conference on Uncertainty in Artificial Intelligence (pp. 116–123). Morgan Kaufmann, San Francisco, CA
Chaudhuri, S., Tan, G. L. (2010). On qualitative comparison of partial regression coefficients for gaussian graphical Markov models. In M. A. G. Viana, H.P. Wynn (Eds.), Algebraic methods in statistics and probability II. Contemporary mathematics, Vol. 516 (pp. 125–133). Providence, Rhode Island: American Mathematical Society.
Cheng, J., Greiner, R., Kelly, J., Bell, D., Liu, W. (2002). Learning bayesian networks from data: an information-theory based approach. Artificial Intelligence, 137, 43–90.
Chickering, D., Meek, C. (2006). On the compatibility of faithfulness and monotone dag faithfulness. Artificial Intelligence, 170, 653–666.
Cover, T., Thomas, J. (2006). Elements of Information Theory. Hoboken: Wiley
Drton, M., Strumfels, B., Sullivant, S. (2007). Algebraic factor analysis: tetrads, pentads and beyond. Probability and related fields, 138, 463–493.
Greenland, S. (2003). Quantifying biases in causal models: classical confounding versus collider-stratification bias. Epidemiology, 14, 300–306.
Greenland, S., Pearl, J. (2011). Adjustments and their consequences-collapsibility analysis using graphical models. International Statistical Review, 79(3), 401–426.
Jarvie, H. P., Neal, C., Withers, P. J., Wescott, C., Acornley, R. M. (2005) Nutrient hydrochemistry for a groundwater-dominated catchment: the Hampshire Avon, UK. Science of The Total Environment, 344 (1–3), 143–158.
Kuroki, M., Cai, Z. (2006). On recovering a population covariance matrix in the presence of selection bias. Biometrika, 93(3), 601–611.
Lauritzen, S. (1996). Graphical models. Oxford: Oxford University Press, Inc.
Lněnička, R., Matúš, F. (2007). On Gaussian conditional independence structures. Kybernetika, 43(3), 327–342.
Matúš, F. (2005). Conditional independence in gaussian vectors and rings of polynomials. In G. Kern-Isberner, W. Rödder, F. Kulmann (Eds.), Conditionals, information, and inference (WCII 2002 Hagen) (pp. 152–161). Berlin: Springer.
Matúš, F. (2006). Piecewise linear conditional information inequality. IEEE Transaction on Information Theory, 52(1), 236–238.
Matúš, F. (2007). Infinitely many information inequalities. In IEEE international symposium on information theory, 2007. ISIT 2007 (pp. 41–44).
Richardson, T., Spirtes, P. (2002). Ancestral graph markov models. The Annals of Statistics, 30(4), 962–1030.
Roberts, G. O., Sahu, S. K. (1997). Updating schemes, correlation structure, blocking and parameterization for the Gibbs sampler. Journal of the Royal Statistical Society Series B Methodological, 59(2), 291–317.
Rodríguez-Iturbe, I., Rinaldo, A. (2001). Fractal River Basins chance and self-organisation. Cambridge: Cambridge University Press.
Shimizu, S., Hoyer, P. O., Hyvärinen, A., Kerminen, A. (2006). A linear non-gaussian acyclic model for causal discovery. Journal of Machine Learning Research, 7, 2003–2030.
Spirtes, P., Glymour, C., Scheines, R. (2000). Causation, prediction, and search. Cambridge: MIT Press.
VanderWeele, T. J., Robins, J. M. (2007). Directed acyclic graphs, sufficient causes, and the properties of confounding on a common effect. American Journal of Epidemiology, 166(9), 1096–1104.
VanderWeele, T. J., Robins, J. M. (2010). Signed directed acyclic graphs for causal inference. Journal of the Royal Statistical Society: Series B Methodological, 72(1), 111–127.
Verma, T., Pearl, J. (1990). Equivalence and synthesis of causal models. In P. Bonissone, M. Henrion, L. Kanal, J. Lemmer (Eds.), Proceedings of the Sixth Conference Conference on Uncertainty in Artificial Intelligence (pp. 220–227). Corvallis: AUAI Press.
Wermuth, N., Cox, D. R. (2008). Distortion of effects caused by indirect confounding. Biometrika, 95(1), 17–33.
Whittaker, J. (2008). Graphical models in applied multivariate statistics. Chichester: Wiley.
Xu, L., Pearl, J. (1989). Structuring causal tree models with continuous variables. In M. Henrion, R. Shachter, L. Kanal, J. Lemmer (Eds.), Proceedings of the fifth conference on uncertainty in artificial intelligence (pp. 170–178). Corvallis: AUAI Press.
Zhang, Z., Yeung, R. W. (1997). A non-shannon-type conditional inequality of information quantities. IEEE Transaction on Information Theory, 43(6), 1982–1986.
Acknowledgments
The author would like to thank Michael Perlman, Thomas Richardson, Mathias Drton, Antar Bandyopadhyay, the referees and the associate editor for their useful comments and suggestions during the preparation of this article.
Author information
Authors and Affiliations
Corresponding author
Additional information
This research was partially supported by Grant R-155-000-081-112 from National University of Singapore.
Electronic supplementary material
Below is the link to the electronic supplementary material.
About this article
Cite this article
Chaudhuri, S. Qualitative inequalities for squared partial correlations of a Gaussian random vector. Ann Inst Stat Math 66, 345–367 (2014). https://doi.org/10.1007/s10463-013-0417-x
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10463-013-0417-x