Advertisement

Probabilistic treatment of the uncertainty from the finite size of weighted Monte Carlo data

Regular Article

Abstract.

Parameter estimation in HEP experiments often involves Monte Carlo simulation to model the experimental response function. A typical application are forward-folding likelihood analyses with re-weighting, or time-consuming minimization schemes with a new simulation set for each parameter value. Problematically, the finite size of such Monte Carlo samples carries intrinsic uncertainty that can lead to a substantial bias in parameter estimation if it is neglected and the sample size is small. We introduce a probabilistic treatment of this problem by replacing the usual likelihood functions with novel generalized probability distributions that incorporate the finite statistics via suitable marginalization. These new PDFs are analytic, and can be used to replace the Poisson, multinomial, and sample-based unbinned likelihoods, which covers many use cases in high-energy physics. In the limit of infinite statistics, they reduce to the respective standard probability distributions. In the general case of arbitrary Monte Carlo weights, the expressions involve the fourth Lauricella function \(F_{D}\), for which we find a new finite-sum representation in a certain parameter setting. The result also represents an exact form for Carlson’s Dirichlet average \(R_{n}\) with \(n > 0\), and thereby an efficient way to calculate the probability generating function of the Dirichlet-multinomial distribution, the extended divided difference of a monomial, or arbitrary moments of univariate B-splines. We demonstrate the bias reduction of our approach with a typical toy Monte Carlo problem, estimating the normalization of a peak in a falling energy spectrum, and compare the results with previously published methods from the literature.

References

  1. 1.
    Louis Lyons, Contemp. Phys. 54, 1 (2013)ADSCrossRefGoogle Scholar
  2. 2.
    F. Halzen, Eur. Phys. J. C 46, 669 (2006)ADSCrossRefGoogle Scholar
  3. 3.
    Lyndon Evans, Philip Bryant, JINST 3, S08001 (2008)Google Scholar
  4. 4.
    Kurt Binder, Dieter W. Heermann, Monte Carlo simulation in statistical physics, in Graduate Texts in Physics (Springer, 2010)Google Scholar
  5. 5.
    Roger Barlow, Christine Beeston, Comput. Phys. Commun. 77, 219 (1993)ADSCrossRefGoogle Scholar
  6. 6.
    Dmitry Chirkin, Likelihood description for comparing data with simulation of limited statistics (2016) available on-line at http://inspirehep.net/record/1413017?ln=de
  7. 7.
    G. Bohm, G. Zech, Nucl. Instrum. Methods Phys. Res. Sect. A 691, 171 (2012)ADSCrossRefGoogle Scholar
  8. 8.
    G. Bohm, G. Zech, Nucl. Instrum. Methods Phys. Res. Sect. A 748, 1 (2014)ADSCrossRefGoogle Scholar
  9. 9.
    F. Beaujean, H.C. Eggers, W.E. Kerzendorf, Mon. Not. R. Astron. Soc. 477, 3425 (2018)ADSCrossRefGoogle Scholar
  10. 10.
    Ritu Aggarwal, Allen Caldwell, Eur. Phys. J. Plus 127, 24 (2012)CrossRefGoogle Scholar
  11. 11.
    C.H. Sim, Stat. Prob. Lett. 15, 135 (1992)CrossRefGoogle Scholar
  12. 12.
    F. Di Salvo, G. Lovison, Parametric inference on samples with random weights, unpublished (1998)Google Scholar
  13. 13.
    F. Di Salvo, Integral Transforms Spec. Funct. 19, 563 (2008)MathSciNetCrossRefGoogle Scholar
  14. 14.
    P.G. Moschopoulos, Ann. Inst. Stat. Math. 37, 541 (1985)MathSciNetCrossRefGoogle Scholar
  15. 15.
    Harold Exton, Multiple hypergeometric functions and applications (Horwood, 1976)Google Scholar
  16. 16.
    James M. Dickey, J. Am. Stat. Assoc. 78, 628 (1983)CrossRefGoogle Scholar
  17. 17.
    Willard Miller, J. Math. Phys. 13, 1393 (1972)ADSCrossRefGoogle Scholar
  18. 18.
    Akio Hattori, Tosihusa Kimura, J. Math. Soc. Jpn. 26, 1 (1974)ADSCrossRefGoogle Scholar
  19. 19.
    B.C. Carlson, J. Math. Anal. Appl. 7, 452 (1963)MathSciNetCrossRefGoogle Scholar
  20. 20.
    Bille Chandler Carlson, Special Functions of Applied Mathematics (Academic Press, 1977)Google Scholar
  21. 21.
    Edward Neuman, Patrick J. Van Fleet, J. Comput. Appl. Math. 53, 225 (1994)MathSciNetCrossRefGoogle Scholar
  22. 22.
    N. Balakrishnan, Discrete Multivariate Distributions (Wiley Online Library, 1997)Google Scholar
  23. 23.
    Ping Zhou, J. Comput. Appl. Math. 236, 94 (2011)MathSciNetCrossRefGoogle Scholar
  24. 24.
    Youneng Ma, Jinhua Yu, Yuanyuan Wang, J. Appl. Math. 2014, 895036 (2014)Google Scholar
  25. 25.
    J.D. Keckic Mitrinovic, S. Dragoslav, The Cauchy Method of Residues (Springer Netherlands, 1984)Google Scholar
  26. 26.
    B.C. Carlson, J. Approx. Theory 67, 311 (1991)MathSciNetCrossRefGoogle Scholar
  27. 27.
    G.S. Monti, The shifted-scaled Dirichlet distribution in the simplex, Universitat de Girona, Departament d’Informàtica i Matemàtica Aplicada (2011)Google Scholar
  28. 28.
    Jieqing Tan, Ping Zhou, Adv. Comput. Math. 23, 333 (2005)MathSciNetCrossRefGoogle Scholar
  29. 29.
    Norman Lloyd Johnson, Samuel Kotz, Urn models and their applicationGoogle Scholar
  30. 30.
    Harold Jeffreys, Proc. R. Soc. London A 186, 453 (1946)CrossRefGoogle Scholar
  31. 31.
    M.G. Aartsen, M. Ackermann, J. Adams, J.A. Aguilar, M. Ahlers, M. Ahrens, D. Altmann, T. Anderson, C. Arguelles, T.C. Arlen et al., Astrophys. J. 796, 109 (2014)ADSCrossRefGoogle Scholar
  32. 32.
    Ilía Nikolaevich Bronshtein and Konstantin Adol’fovich Semendyayev. Handbook of Mathematics (Springer Science & Business Media, 2013)Google Scholar
  33. 33.
    Peter J.M. van Laarhoven, Ton A.C.M. Kalker, J. Comput. Appl. Math. 21, 369 (1988)MathSciNetCrossRefGoogle Scholar
  34. 34.
    Georgy P. Egorychev, Integral Representation and the Computation of Combinatorial Sums, in Mathematical Monographs, Vol. 59 (American Mathematical Soc., 1984)Google Scholar

Copyright information

© Società Italiana di Fisica and Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Erlangen Centre for Astroparticle Physics (ECAP)ErlangenGermany

Personalised recommendations