Abstract
Additive Bayesian networks (ABN) are types of graphical models that extend the usual Bayesian-generalised linear model to multiple dependent variables through the factorisation of the joint probability distribution of the underlying variables. When fitting an ABN model, the choice of the prior for the parameters is of crucial importance. If an inadequate prior—like a not sufficiently informative one—is used, data separation and data sparsity may lead to issues in the model selection process. In this work we present a simulation study to compare two weakly informative priors with a strongly informative one. For the weakly informative prior, we use a zero mean Gaussian prior with a large variance, currently implemented in the R package abn. The candidate prior belongs to the Student’s t-distribution. It is specifically designed for logistic regressions. Finally, the strongly informative prior is Gaussian with a mean equal to the true parameter value and a small variance. We compare the impact of these priors on the accuracy of the learned additive Bayesian network as function of different parameters. We create a simulation study to illustrate Lindley’s paradox based on the prior choice. We then conclude by highlighting the good performance of the informative Student’s t-prior and the limited impact of Lindley’s paradox. Finally, suggestions for further developments are provided.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Chen, M., Ibrahim, J.G.: Conjugate priors for generalized linear models. Statistica Sinica 13, 461–476 (2003)
Diaconis, P., Ylvisaker, D.: Conjugate priors for exponential families. Ann. Stat. 7(2), 269–281 (1979)
Djebbari, A., Quackenbush, J.: Seeded Bayesian networks: constructing genetic networks from microarray data. BMC Syst. Biol. 2(1), 57 (2008)
Dojer, N., Gambin, A., Mizera, A., Wilczyński, B., Tiuryn, J.: Applying dynamic Bayesian networks to perturbed gene expression data. BMC Bioinform. 7(1), 249 (2006)
Firth, D.: Bias reduction of maximum likelihood estimates. Biometrika 80(1), 27–38 (1993)
Flesch, I., Lucas, P.J.: Markov equivalence in Bayesian networks. In: Lucas, P., Gámez, J.A., Salmerón, A. (eds.) Advances in Probabilistic Graphical Models, pp. 3–38. Springer, Berlin, Heidelberg (2007)
Gelman, A., Stern, H.S., Carlin, J.B., Dunson, D.B., Vehtari, A., Rubin, D.B.: Bayesian Data Analysis. Chapman and Hall/CRC (2013)
Gelman, A., Jakulin, A., Pittau, M.G., Su, Y.S.: A weakly informative default prior distribution for logistic and other regression models. Ann. Appl. Stat. 2(4), 1360–1383 (2008)
Gutiérrez-Peña, E., Smith, A.F.M.: Conjugate parameterizations for natural exponential families. J. Am. Stat. Assoc. 90(432), 1347–1356 (1995)
Hartnack, S., Springer, S., Pittavino, M., Grimm, H.: Attitudes of Austrian veterinarians towards euthanasia in small animal practice: impacts of age and gender on views on euthanasia. BMC Vet. Res. 12(1), 26 (2016)
Heckerman, D., Geiger, D., Chickering, D.M.: Learning Bayesian networks: the combination of knowledge and statistical data. Mach. Learn. 20(3), 197–243 (1995)
Hodges, A.P., Dai, D., Xiang, Z., Woolf, P., Xi, C., He, Y.: Bayesian network expansion identifies new ROS and biofilm regulators. PLOS One 5(3), e9513 (2010)
Jansen, R., Yu, H., Greenbaum, D., Kluger, Y., Krogan, N.J., Chung, S., Emili, A., Snyder, M., Greenblatt, J.F., Gerstein, M.: A Bayesian networks approach for predicting protein-protein interactions from genomic data. Science 302(5644), 449–453 (2003)
Koivisto, M., Sood, K.: Exact Bayesian structure discovery in Bayesian networks. J. Mach. Learn. Res. 5(May), 549–573 (2004)
Kratzer, G., Pittavino, M., Ian, L.F., Lewis, I.F.: abn: an R package for modelling multivariate data using additive Bayesian networks, R package version 1.3 (2018). https://CRAN.R-project.org/package=abn
Kratzer G, Furrer R (2018) Information-Theoretic Scoring Rules to Learn Additive Bayesian Network Applied to Epidemiology. arXiv:1808.01126
Lewis, F.I.: Bayesian networks as a tool for epidemiological systems analysis. In: AIP Conference Proceedings vol. 1493, pp. 610–617 (2012)
Lewis, F.I., Brülisauer, F., Gunn, G.J.: Structure discovery in Bayesian networks: an analytical tool for analysing complex animal health data. Prev. Vet. Med. 100(2), 109–115 (2011)
Lewis, F.I., McCormick, B.J.: Revealing the complexity of health determinants in resource-poor settings. Am. J. Epidemiol. 176(11), 1051–1059 (2012)
Lewis, F.I., Ward, M.P.: Improving epidemiologic data analyses through multivariate regression modelling. Emerg. Themes Epidemiol. 10(1), 4 (2013)
Lindley, D.V.: A statistical paradox. Biometrika 44(1/2), 187–192 (1957)
Pitman, E.J.G.: Sufficient statistics and intrinsic accuracy. Math. Proc. Camb. Philos. Soc. 32(4), 567–579 (1936)
Pittavino, M.: Additive Bayesian networks for multivariate data: parameter learning, model fitting and applications in veterinary epidemiology. Ph.D. thesis, University of Zurich (2016)
Pittavino, M., Dreyfus, A., Heuer, C., Benschop, J., Wilson, P., Collins-Emerson, J., Torgerson, P.R., Furrer, R.: Comparison between generalized linear modelling and additive Bayesian network; identification of factors associated with the incidence of antibodies against Leptospira interrogans sv Pomona in meat workers in New Zealand. Acta Trop. 173, 191–199 (2017)
Poon, A.F.Y., Lewis, F.I., Pond, S.L.K., Frost, S.D.W.: Evolutionary interactions between N-linked glycosylation sites in the HIV-1 envelope. PLOS Comput. Biol. 3(1), e11 (2007)
R Core Team: R: a Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria (2017)
Robinson, R.W.: Counting unlabeled acyclic digraphs. In: Little, C.H.C. (ed.) Combinatorial Mathematics V, pp. 28–43. Springer, Berlin, Heidelberg (1977)
Sanchez-Vazquez, M.J., Nielen, M., Edwards, S.A., Gunn, G.J., Lewis, F.I.: Identifying associations between pig pathologies using a multi-dimensional machine learning methodology. BMC Vet. Res. 8(1), 151 (2012)
Ward, M.P., Lewis, F.I.: Bayesian graphical modelling: applications in veterinary epidemiology. Prev. Vet. Med. 110(1), 1–3 (2013)
Zorn, C.: A solution to separation in binary response models. Polit. Anal. 13(2), 157–170 (2005)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this paper
Cite this paper
Kratzer, G., Furrer, R., Pittavino, M. (2019). Comparison Between Suitable Priors for Additive Bayesian Networks. In: Argiento, R., Durante, D., Wade, S. (eds) Bayesian Statistics and New Generations. BAYSM 2018. Springer Proceedings in Mathematics & Statistics, vol 296. Springer, Cham. https://doi.org/10.1007/978-3-030-30611-3_10
Download citation
DOI: https://doi.org/10.1007/978-3-030-30611-3_10
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-30610-6
Online ISBN: 978-3-030-30611-3
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)