Skip to main content

Absolute continuity and singularity of distributions of dependent observations: Gaussian and exchangeable measures

Research Articles

Part of the Lecture Notes in Mathematics book series (LNM,volume 1210)

Abstract

A general version of Kakutani's dichotomy theorem concerning absolute continuity and singularity of measures on infinite product spaces is proved. Based on this result it is possible to deduce certain dichotomy properties for product measures, Gaussian measures and exchangeable measures.

On the other hand the same arguments can be used to characterize those exchangeable distributions which are presentable as a mixture of product measures in de Finetti's sense. Furthermore it is proved that each exchangeable probability measure can be decomposed in a part which is presentable and a second part being mutually singular with respect to all presentable distributions.

Keywords

  • Probability Measure
  • Product Measure
  • Gaussian Measure
  • Absolute Continuity
  • Dependent Random Variable

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

This is a preview of subscription content, access via your institution.

Buying options

Chapter
USD   29.95
Price excludes VAT (Canada)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD   34.99
Price excludes VAT (Canada)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD   46.00
Price excludes VAT (Canada)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Learn about institutional subscriptions

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Blum, J.R. (1982). Exchangeability and Quasi-Exchangeability. In: Exchangeability in Prob. and Statist., G. Koch and F. Spizzichino (editors), North-Holland, 171–176.

    Google Scholar 

  2. Breiman, L., LeCam, L. Schwartz, L. (1964). Consistent estimates and zero-one sets. Ann. Math. Statistics 35, 157–161

    CrossRef  MathSciNet  MATH  Google Scholar 

  3. Chatterji, S.D., Mandrekar, V. (1978). Equivalence and singularity of Gaussian measures and applications. In: Probabilistic analysis and related topics, 169–195. Edited by A.T. Bharucha-Reid.

    Google Scholar 

  4. Chatterji, S.D., Ramaswamy, S. (1982). Mesures Gaussiennes et mesures produit. Lect. Notes Math. 920, 570–580. Springer.

    CrossRef  MathSciNet  MATH  Google Scholar 

  5. Dubins, L., Freedman, D. (1964). Measurable sets of measures. Pacific J. Math. 14. 1211–1222.

    CrossRef  MathSciNet  MATH  Google Scholar 

  6. Dubins, L. Freedman, D. (1979). Exchangeable processes need not be mixtures of independent identically distributed random variables. Zeitschr. Wahrscheinlichkeitstheorie verw. Geb. 48, 115–132.

    CrossRef  MathSciNet  MATH  Google Scholar 

  7. Dubins, L. (1983). Some exchangeable probabilities are singular with respect to all representable probabilities. Zeitschr. Wahrscheinlichkeitstheorie verw. Geb. 64, 1–5.

    CrossRef  MathSciNet  MATH  Google Scholar 

  8. Engelbert, H.J., Sirjaev, A.N. (1980). On absolute continuity and singularity of probability measure. Mathem. Statistics. Banach Center public., Vol. 6. PWN-Polish scientific publishers, Warsaw.

    Google Scholar 

  9. Fernique, X. (1984). Comparaison de mesures Gaussiennes et de mesure produit. Ann. Inst. Henri Poincaré, Prob. Statist. 20, 165–175.

    MathSciNet  MATH  Google Scholar 

  10. Fernique, X. (1985). Comparison de mesures Gaussiennes et de mesures produit dans les espaces de Fréchet separables. Lecture Notes Math. 1153, 179–197, Springer Verlag.

    CrossRef  MathSciNet  MATH  Google Scholar 

  11. Freedman, D. (1980). A mixture of independent identically distributed random variables need not admit a regular conditional probability given the exchangeable 6-field. Zeitschr. Wahrscheinlichkeitstheorie 51, 239–248.

    CrossRef  MathSciNet  MATH  Google Scholar 

  12. Janssen, A. (1981). Meßbare Mengen von Maßen und Inhalten. Manuscripta Math. 34, 1–15.

    CrossRef  MathSciNet  MATH  Google Scholar 

  13. Janssen, A. (1982). Zero-one laws for infinitely divisible probability measures on groups. Zeitschr. Wahrscheinlichkeitstheorie 60, 119–138.

    MathSciNet  MATH  Google Scholar 

  14. Kabanov, J.M., Lipcer, R.S., Sirjaev, A.N. (1977). On the question of absolute continuity and singularity of probability measures. Math. USSR Sbornik 33, 203–221.

    CrossRef  MathSciNet  Google Scholar 

  15. Kakutani, S. (1948). On equivalence of infinite product measures. Ann. of Math. 49, 214–224.

    CrossRef  MathSciNet  MATH  Google Scholar 

  16. LeCam, L. (1969). Théorie asymptotique de la décision statistique. Les presses de l'université de Montréal, Canada.

    Google Scholar 

  17. Ramaswamy, S. (1983). Gaussian measures and product measures, Preprint.

    Google Scholar 

  18. Skorohod, A.V. (1974). Integration in Hilbert space. Ergebn. d. Mathem. 79, Springer.

    Google Scholar 

  19. Strasser, H. (1985): Mathematical theory of Statistics, de Gruyter Studies in Mathematics 7.

    Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Editor information

Editors and Affiliations

Rights and permissions

Reprints and Permissions

Copyright information

© 1986 Springer-Verlag

About this paper

Cite this paper

Janssen, A. (1986). Absolute continuity and singularity of distributions of dependent observations: Gaussian and exchangeable measures. In: Heyer, H. (eds) Probability Measures on Groups VIII. Lecture Notes in Mathematics, vol 1210. Springer, Berlin, Heidelberg. https://doi.org/10.1007/BFb0077177

Download citation

  • DOI: https://doi.org/10.1007/BFb0077177

  • Published:

  • Publisher Name: Springer, Berlin, Heidelberg

  • Print ISBN: 978-3-540-16806-5

  • Online ISBN: 978-3-540-44852-5

  • eBook Packages: Springer Book Archive