Skip to main content

2020 | OriginalPaper | Buchkapitel

Efficient Stochastic Approaches for Multidimensional Integrals in Bayesian Statistics

verfasst von : Venelin Todorov, Ivan Dimov

Erschienen in: Large-Scale Scientific Computing

Verlag: Springer International Publishing

Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.

search-config
loading …

Abstract

A fundamental problem in Bayesian statistics is the accurate evaluation of multidimensional integrals. A comprehensive experimental study of quasi-Monte Carlo algorithms based on Sobol sequence combined with Matousek linear scrambling and a comparison with adaptive Monte Carlo approach and a lattice rule based on generalized Fibonacci numbers has been presented. The numerical tests show that the stochastic algorithms under consideration are efficient for multidimensional integration and especially for computing high dimensional integrals. It is a crucial element since this may be important to be estimated in order to achieve a more accurate and reliable interpretation of the results in Bayesian statistics which is foundational in applications such as machine learning.

Sie haben noch keine Lizenz? Dann Informieren Sie sich jetzt über unsere Produkte:

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

  • über 102.000 Bücher
  • über 537 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Maschinenbau + Werkstoffe
  • Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Springer Professional "Technik"

Online-Abonnement

Mit Springer Professional "Technik" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 390 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Maschinenbau + Werkstoffe




 

Jetzt Wissensvorsprung sichern!

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 340 Zeitschriften

aus folgenden Fachgebieten:

  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Versicherung + Risiko




Jetzt Wissensvorsprung sichern!

Literatur
1.
Zurück zum Zitat Antonov, I., Saleev, V.: An economic method of computing LP tau-sequences. USSR Comput. Math. Math. Phys. 19, 252–256 (1980)CrossRef Antonov, I., Saleev, V.: An economic method of computing LP tau-sequences. USSR Comput. Math. Math. Phys. 19, 252–256 (1980)CrossRef
3.
Zurück zum Zitat Bendavid, J.: Efficient Monte Carlo integration using boosted decision trees and generative deep neural networks. arXiv preprint arXiv:1707.00028 (2017) Bendavid, J.: Efficient Monte Carlo integration using boosted decision trees and generative deep neural networks. arXiv preprint arXiv:​1707.​00028 (2017)
4.
Zurück zum Zitat Boyle, P., Lai, Y., Tan, K.: Using lattice rules to value low-dimensional derivative contracts (2001) Boyle, P., Lai, Y., Tan, K.: Using lattice rules to value low-dimensional derivative contracts (2001)
5.
Zurück zum Zitat Dimov, I.: Monte Carlo Methods for Applied Scientists, 291 p. World Scientific, Singapore (2008)MATH Dimov, I.: Monte Carlo Methods for Applied Scientists, 291 p. World Scientific, Singapore (2008)MATH
6.
Zurück zum Zitat Dimov, I., Georgieva, R.: Monte Carlo algorithms for evaluating Sobol’ sensitivity indices. Math. Comput. Simul. 81(3), 506–514 (2010)MathSciNetCrossRef Dimov, I., Georgieva, R.: Monte Carlo algorithms for evaluating Sobol’ sensitivity indices. Math. Comput. Simul. 81(3), 506–514 (2010)MathSciNetCrossRef
7.
Zurück zum Zitat Dimov, I., Karaivanova, A.: Error analysis of an adaptive Monte Carlo method for numerical integration. Math. Comput. Simul. 47, 201–213 (1998)MathSciNetCrossRef Dimov, I., Karaivanova, A.: Error analysis of an adaptive Monte Carlo method for numerical integration. Math. Comput. Simul. 47, 201–213 (1998)MathSciNetCrossRef
8.
Zurück zum Zitat Hua, L.K., Wang, Y.: Applications of Number Theory to Numerical analysis. Springer, Heidelberg (1981)MATH Hua, L.K., Wang, Y.: Applications of Number Theory to Numerical analysis. Springer, Heidelberg (1981)MATH
9.
Zurück zum Zitat Hong, H.S., Hickernell, F.J.: ALGORITHM 823: implementing scrambled digital sequences. ACM Trans. Math. Softw. 29(2), 95–109 (2003)MathSciNetCrossRef Hong, H.S., Hickernell, F.J.: ALGORITHM 823: implementing scrambled digital sequences. ACM Trans. Math. Softw. 29(2), 95–109 (2003)MathSciNetCrossRef
10.
Zurück zum Zitat Korobov, N.M.: Number-Theoretical Methods in Approximate Analysis. Fizmatgiz, Moscow (1963) Korobov, N.M.: Number-Theoretical Methods in Approximate Analysis. Fizmatgiz, Moscow (1963)
11.
Zurück zum Zitat Lin, S.: Algebraic methods for evaluating integrals in Bayesian statistics. Ph.D. dissertation, UC Berkeley, May 2011 Lin, S.: Algebraic methods for evaluating integrals in Bayesian statistics. Ph.D. dissertation, UC Berkeley, May 2011
12.
Zurück zum Zitat Lin, S., Sturmfels, B., Xu, Z.: Marginal likelihood integrals for mixtures of independence models. J. Mach. Learn. Res. 10, 1611–1631 (2009)MathSciNetMATH Lin, S., Sturmfels, B., Xu, Z.: Marginal likelihood integrals for mixtures of independence models. J. Mach. Learn. Res. 10, 1611–1631 (2009)MathSciNetMATH
13.
Zurück zum Zitat Matousek, J.: On the L2-discrepancy for anchored boxes. J. Complex. 14(4), 527–556 (1998)CrossRef Matousek, J.: On the L2-discrepancy for anchored boxes. J. Complex. 14(4), 527–556 (1998)CrossRef
14.
Zurück zum Zitat Owen, A.: Variance and discrepancy with alternative scramblings. ACM Trans. Comput. Logic 13, 1–16 (2002) Owen, A.: Variance and discrepancy with alternative scramblings. ACM Trans. Comput. Logic 13, 1–16 (2002)
15.
Zurück zum Zitat Paskov, S.: Computing high dimensional integrals with applications to finance. Columbia University (1994, preprint) Paskov, S.: Computing high dimensional integrals with applications to finance. Columbia University (1994, preprint)
16.
Zurück zum Zitat Sloan, I.H., Kachoyan, P.J.: Lattice methods for multiple integration: theory, error analysis and examples. SIAM J. Numer. Anal. 24, 116–128 (1987)MathSciNetCrossRef Sloan, I.H., Kachoyan, P.J.: Lattice methods for multiple integration: theory, error analysis and examples. SIAM J. Numer. Anal. 24, 116–128 (1987)MathSciNetCrossRef
17.
Zurück zum Zitat Song, J., Zhao, S., Ermon, S.: A-NICE-MC: adversarial training for MCMC. In: Advances in Neural Information Processing Systems, pp. 5140–5150 (2017) Song, J., Zhao, S., Ermon, S.: A-NICE-MC: adversarial training for MCMC. In: Advances in Neural Information Processing Systems, pp. 5140–5150 (2017)
19.
Zurück zum Zitat Watanabe, S.: Algebraic analysis for nonidentifiable learning machines. Neural Comput. 13, 899–933 (2001)CrossRef Watanabe, S.: Algebraic analysis for nonidentifiable learning machines. Neural Comput. 13, 899–933 (2001)CrossRef
Metadaten
Titel
Efficient Stochastic Approaches for Multidimensional Integrals in Bayesian Statistics
verfasst von
Venelin Todorov
Ivan Dimov
Copyright-Jahr
2020
DOI
https://doi.org/10.1007/978-3-030-41032-2_52