Advertisement

Probability Theory and Related Fields

, Volume 168, Issue 3–4, pp 901–917 | Cite as

Existence and consistency of Wasserstein barycenters

  • Thibaut Le Gouic
  • Jean-Michel Loubes
Article

Abstract

Based on the Fréchet mean, we define a notion of barycenter corresponding to a usual notion of statistical mean. We prove the existence of Wasserstein barycenters of random probabilities defined on a geodesic space (Ed). We also prove the consistency of this barycenter in a general setting, that includes taking barycenters of empirical versions of the probability measures or of a growing set of probability measures.

Keywords

Fréchet mean Wasserstein distance Barycenter Existence Consistency 

Mathematics Subject Classification

60D05 62H10 62G20 

References

  1. 1.
    Agueh, M., Carlier, G.: Barycenters in the wasserstein space. SIAM J. Math. Anal. 43(2), 904–924 (2011)MathSciNetCrossRefzbMATHGoogle Scholar
  2. 2.
    Agulló-Antolín, M., Cuesta-Albertos, J.A., Lescornel, H., Loubes, J.-M.: A parametric registration model for warped distributions with Wasserstein’s distance. J. Multivar. Anal. 135, 117–130 (2015)MathSciNetCrossRefzbMATHGoogle Scholar
  3. 3.
    Álvarez-Esteban, P.C., del Barrio, E., Cuesta-Albertos, J. A., Matrán, C.: Wide consensus for parallelized inference (2015). arXiv:1511.05350
  4. 4.
    Ambrosio, L., Gigli, N., Savaré, G.: Gradient flows: in metric spaces and in the space of probability measures. Springer Science & Business Media (2008)Google Scholar
  5. 5.
    Amit, Y., Grenander, U., Piccioni, M.: Structural image restoration through deformable template. J. Am. Stat. Assoc. 86, 376–387 (1991)CrossRefGoogle Scholar
  6. 6.
    Baum, M., Willett, P., Hanebeck, U.D.: On wasserstein barycenters and mmospa estimation. Signal Process. Lett. IEEE 22(10), 1511–1515 (2015)CrossRefGoogle Scholar
  7. 7.
    Bercu, B., Fraysse, P.: A Robbins-Monro procedure for estimation in semiparametric regression models. Ann. Stat. 40(2), 666–693 (2012)MathSciNetCrossRefzbMATHGoogle Scholar
  8. 8.
    Berti, P., Pratelli, L., Rigo, P., et al.: Gluing lemmas and Skorohod representations. Electron. Commun. Probab. 20 (2015)Google Scholar
  9. 9.
    Bigot J., Klein, T.: Consistent estimation of a population barycenter in the Wasserstein space (2012). arXiv:1212.2562
  10. 10.
    Bishop, A.N.: Information fusion via the wasserstein barycenter in the space of probability measures: direct fusion of empirical measures and gaussian fusion with unknown correlation. In: Information Fusion (FUSION), 2014 17th International Conference on, pp. 1–7. IEEE (2014)Google Scholar
  11. 11.
    Boissard, E., Le Gouic, T., Loubes, J.-M.: Distribution’s template estimate with Wasserstein metrics. Bernoulli 21(2), 740–759 (2015)MathSciNetCrossRefzbMATHGoogle Scholar
  12. 12.
    Bolstad, B.M., Irizarry, R.A., Åstrand, M., Speed, T.P.: A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics 19(2), 185–193 (2003)CrossRefGoogle Scholar
  13. 13.
    Brown, L.D., Purves, R., et al.: Measurable selections of extrema. Ann. Stat. 1(5), 902–912 (1973)MathSciNetCrossRefzbMATHGoogle Scholar
  14. 14.
    Burago, D., Burago, Y., Sergei, I.: A Course in Metric Geometry, vol. 33, pp. 371–374. American Mathematical Society, Providence (2001)Google Scholar
  15. 15.
    Czado, C., Munk, A.: Assessing the similarity of distributions–finite sample performance of the empirical Mallows distance. J. Stat. Comput. Simul. 60(4), 319–346 (1998)MathSciNetCrossRefzbMATHGoogle Scholar
  16. 16.
    Del Barrio, E., Lescornel, H., Loubes, J.-M.: A statistical analysis of a deformation model with Wasserstein barycenters: estimation procedure and goodness of fit test. (2015). arXiv:1508.06465
  17. 17.
    Gallon, S., Loubes, J.-M., Maza, E.: Statistical properties of the quantile normalization method for density curve alignment. Math. Biosci. 242(2), 129–142 (2013)MathSciNetCrossRefzbMATHGoogle Scholar
  18. 18.
    Gamboa, F., Loubes, J.-M., Maza, E.: Semi-parametric estimation of shifts. Electron. J. Stat. 1, 616–640 (2007)MathSciNetCrossRefzbMATHGoogle Scholar
  19. 19.
    Grenander, U.: General pattern theory—a mathematical study of regular structures. Clarendon Press, Oxford (1993)zbMATHGoogle Scholar
  20. 20.
    Kendall, D.G., Barden, D., Carne, T.K., Le, H.: Shape and shape theory, Wiley Series in Probability and Statistics. Wiley, Chichester (1999)CrossRefzbMATHGoogle Scholar
  21. 21.
    Kim, Y.-H., Pass, B.: Wasserstein Barycenters over Riemannian manfolds (2014). arXiv:1412.7726
  22. 22.
    Le Gouic, T.: Localisation de masse et espaces de Wasserstein. PhD thesis, Université de Toulouse, Université de Toulouse III-Paul Sabatier (2013)Google Scholar
  23. 23.
    Lott, J., Villani, C.: Ricci curvature for metric-measure spaces via optimal transport. Ann. Math. 196(3), 903–991 (2009)Google Scholar
  24. 24.
    Ohta, S.-I.: Barycenters in alexandrov spaces of curvature bounded below. Adv. Geom. 14, 571–587 (2012)MathSciNetCrossRefzbMATHGoogle Scholar
  25. 25.
    Ramsay, J.O., Silverman, B.W.: Functional data analysis. Springer Series in Statistics, 2nd edn. Springer, New York (2005)Google Scholar
  26. 26.
    Sakoe, H., Chiba, S.: Dynamic programming algorithm optimization for spoken word recognition. IEEE Trans. Acoust. Speech Signal Process. 26(1), 43–49 (1978)CrossRefzbMATHGoogle Scholar
  27. 27.
    Srivastava, S., Li, C., Dunson, D.B.: Scalable Bayes via Barycenter in Wasserstein Space. (2015). arXiv:1508.05880
  28. 28.
    Sturm, K.-T.: Probability measures on metric spaces of nonpositive. In: Heat Kernels and Analysis on Manifolds, Graphs, and Metric Spaces. Lecture Notes from a Quarter Program on Heat Kernels, Random Walks, and Analysis on Manifolds and Graphs: April 16–July 13, 2002, vol. 338, pp. 357. Emile Borel Centre of the Henri Poincaré Institute, Paris (2003)Google Scholar
  29. 29.
    Trouve, A., Younes, L.: Metamorphoses through lie group action. Found. Comput. Math. 5(2), 173–198 (2005)MathSciNetCrossRefzbMATHGoogle Scholar
  30. 30.
    Villani, C: Optimal transport: old and new, vol. 338. Springer Science & Business Media (2008)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2016

Authors and Affiliations

  1. 1.Aix Marseille Université, CNRS, Centrale Marseille, I2M, UMR 7373MarseilleFrance
  2. 2.Université de Toulouse, Institut de Mathématiques de ToulouseToulouseFrance

Personalised recommendations