Inference Given Summary Statistics

  • Habib N. NajmEmail author
  • Kenny Chowdhary
Reference work entry


In many practical situations, where one is interested in employing Bayesian inference methods to infer parameters of interest, a significant challenge is that actual data is not available. Rather, what is most commonly available in the literature are summary statistics on the data, on parameters of interest, or on functions thereof. In this chapter, we present a general framework relying on the maximum entropy principle, and employing approximate Bayesian computation methods, to infer a joint posterior density on parameters of interest given summary statistics, as well as other known details about the experiment or observational system behind the published statistics. By essentially redoing the experimental fitting using proposed data sets, the method ensures that the inferred joint posterior density on model parameters is consistent with the given statistics and with the model.


Approximate bayesian computation Bayesian inference Maximum entropy Missing data Sufficient statistic 


  1. 1.
    Beaumont, M.A., Zhang, W., Balding, D.J.: Approximate Bayesian computation in population genomics. Genetics 162(4), 2025–2035 (2002)Google Scholar
  2. 2.
    Bernardo, J., Smith, A.: Bayesian Theory. Wiley Series in Probability and Statistics. Wiley, Chichester (2000)zbMATHGoogle Scholar
  3. 3.
    Berry, R., Najm, H., Debusschere, B., Adalsteinsson, H., Marzouk, Y.: Data-free inference of the joint distribution of uncertain model parameters. J. Comput. Phys. 231, 2180–2198 (2012)MathSciNetCrossRefzbMATHGoogle Scholar
  4. 4.
    Bevington, P., Robinson, D.: Data Reduction and Error Analysis for the Physical Sciences, 2nd edn. McGraw-Hill, New York (1992)Google Scholar
  5. 5.
    Box, G.E., Hunter, J.S., Hunter, W.G.: Statistics for Experimenters: Design, Innovation, and Discovery, 2nd edn. Wiley, New York (2005)zbMATHGoogle Scholar
  6. 6.
    Carlin, B.P., Louis, T.A.: Bayesian Methods for Data Analysis. Chapman and Hall/CRC, Boca Raton (2011)zbMATHGoogle Scholar
  7. 7.
    Caticha, A., Preuss, R.: Maximum entropy and Bayesian data analysis: entropic prior distributions. Phys. Rev. E 70(4), 046127 (2004)CrossRefGoogle Scholar
  8. 8.
    Chowdhary, K., Najm, H.: Data free inference with processed data products. Stat. Comput. 1–21 (2014). doi:10.1007/s11222-014-9484-yGoogle Scholar
  9. 9.
    Clyde, M.: Bayesian model averaging and model search strategies (with discussion). In: Bernardo, J., Berger, J., Dawid, A., Smith, A. (eds.) Bayesian Statistics 6, pp. 157–185. Oxford University Press, New York (1999)Google Scholar
  10. 10.
    Dupuis, P., Ellis, R.: A Weak Convergence Approach to the Theory of Large Deviations. Wiley-Interscience, New York (1997)CrossRefzbMATHGoogle Scholar
  11. 11.
    Gelman, A., Meng, X.L., Stern, H.: Posterior Predictive Assessment of Model Fitness via Realized Discrepancies. Statistica Sinica 6, 733–807 (1996)MathSciNetzbMATHGoogle Scholar
  12. 12.
    Genest, C.: A characterization theorem for externally Bayesian groups. Ann. Stat. 12(3), 1100–1105 (1984)MathSciNetCrossRefzbMATHGoogle Scholar
  13. 13.
    Genest, C., Zidek, J.: Combining probability distributions: a critique and an annotated bibliography. Stat. Sci. 1(1), 114–135 (1986)MathSciNetCrossRefzbMATHGoogle Scholar
  14. 14.
    Gregory, P.: Bayesian Logical Data Analysis for the Physical Sciences. Cambridge University Press, Cambridge (2010)zbMATHGoogle Scholar
  15. 15.
    Hansen, P.C., Pereyra, V., Scherer, G.: Least Squares Data Fitting with Applications. The Johns Hopkins University Press, Baltimore (2013)zbMATHGoogle Scholar
  16. 16.
    Hebrard, E., Dobrijevic, M.: How measurements of rate coefficients at low temperature increase the predictivity of photochemical models of Titan’s atmosphere. J. Phys. Chem. 113, 11227–11237 (2009)CrossRefGoogle Scholar
  17. 17.
    IJUQ: Reprinted from Najm, H.N., Berry, R.D., Safta, C., Sargsyan, K., Debusschere, B.J.: Data-free inference of uncertain parameters in chemical models. Int. J. Uncertain. Quantif. 4, 111–132 (2014); Copyright (2014); with permission from Begell House, IncGoogle Scholar
  18. 18.
    Jaynes, E., Bretthorst, G.L. (eds.): Probability Theory: The Logic of Science. Cambridge University Press, Cambridge (2003)zbMATHGoogle Scholar
  19. 19.
    Khalil, M., Najm, H.: Probabilistic inference of reaction rate parameters based on summary statistics. In: Proceedings of the 9th U.S. National Combustion Meeting, Cincinnati (2015)Google Scholar
  20. 20.
    Lakowicz, J.R.: Principles of Fluorescence Spectroscopy, 2nd edn. Kluwer Academic, New York (1999). Support Plane Analysis: see pp. 122–123Google Scholar
  21. 21.
    Lehmann, E., Casella, G.: Theory of Point Estimation. Springer Texts in Statistics. Springer, New York (2003). zbMATHGoogle Scholar
  22. 22.
    Lynch, S., Western, B.: Bayesian posterior predictive checks for complex models. Sociol. Methods Res. 32(3), 301–335 (2004). doi:10.1177/0049124103257303MathSciNetCrossRefGoogle Scholar
  23. 23.
    Nagy, T., Turányi, T.: Determination of the uncertainty domain of the arrhenius parameters needed for the investigation of combustion kinetic models. Reliab. Eng. Syst. Saf. 107, 29–34 (2012)CrossRefGoogle Scholar
  24. 24.
    Najm, H., Berry, R., Safta, C., Sargsyan, K., Debusschere, B.: Data free inference of uncertain parameters in chemical models. Int. J. Uncertain. Quantif. 4(2), 111–132 (2014). doi:10.1615/Int.J.UncertaintyQuantification.2013005679MathSciNetCrossRefGoogle Scholar
  25. 25.
    Park, T., Casella, G.: The Bayesian Lasso. J. Am. Stat. Assoc. 103(482), 681–686 (2008)MathSciNetCrossRefzbMATHGoogle Scholar
  26. 26.
    Shannon, C.E.: A mathematical theory of communication. ACM SIGMOBILE Mobile Comput. Commun. Rev. 5(1), 3–55 (2001)MathSciNetCrossRefGoogle Scholar
  27. 27.
    Sisson, S.A., Fan, Y.: Likelihood-free Markov chain Monte Carlo. In: Brooks, S. (ed.) Handbook of Markov Chain Monte Carlo. Chapman & Hall, London (2010)Google Scholar
  28. 28.
    Sivia, D.S., Carlile, C.J.: Molecular-spectroscopy and Bayesian spectral-analysis – how many lines are there. J. Chem. Phys. 96(1), 170 – 178 (1992)CrossRefGoogle Scholar
  29. 29.
    Smith, G., Golden, D., Frenklach, M., Moriarty, N., Eiteneer, B., Goldenberg, M., Bowman, C., Hanson, R., Song, S., Gardiner, W., Jr., Lissianski, V., Zhiwei, Q.: GRI mechanism for methane/air, version 3.0 (1999), 30 July 1999.
  30. 30.
    STCO: Reprinted from the Chowdhary, K., Najm, H.N.: Data free inference with processed data products. J. Stat. Comput. 1–21 (2014); Copyright (2014); with permission from Springer, U.S.Google Scholar
  31. 31.
    Tarantola, A.: Inverse Problem Theory and Methods for Model Parameter Estimation. SIAM, Philadelphia (2005)CrossRefzbMATHGoogle Scholar
  32. 32.
    Tibshirani, R.: Regression shrinkage and selection via the Lasso. J. R. Stat. Soc. Ser. B (Methodol.) 58, 267–288 (1996)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2017

Authors and Affiliations

  1. 1.Combustion Research Facility, Reacting Flow ResearchSandia National LaboratoriesLivermoreUSA
  2. 2.Quantitative Modeling and AnalysisSandia National LaboratoriesLivermoreUSA

Personalised recommendations