The Annals of Statistics

Invariant Bayesian estimation on manifolds

Ian H. Jermyn

Full-text: Open access

Abstract

A frequent and well-founded criticism of the maximum a posteriori (MAP) and minimum mean squared error (MMSE) estimates of a continuous parameter γ taking values in a differentiable manifold Γ is that they are not invariant to arbitrary “reparameterizations” of Γ. This paper clarifies the issues surrounding this problem, by pointing out the difference between coordinate invariance, which is a sine qua non for a mathematically well-defined problem, and diffeomorphism invariance, which is a substantial issue, and then provides a solution. We first show that the presence of a metric structure on Γ can be used to define coordinate-invariant MAP and MMSE estimates, and we argue that this is the natural way to proceed. We then discuss the choice of a metric structure on Γ. By imposing an invariance criterion natural within a Bayesian framework, we show that this choice is essentially unique. It does not necessarily correspond to a choice of coordinates. In cases of complete prior ignorance, when Jeffreys’ prior is used, the invariant MAP estimate reduces to the maximum likelihood estimate. The invariant MAP estimate coincides with the minimum message length (MML) estimate, but no discretization or approximation is used in its derivation.

Article information

Source
Ann. Statist. Volume 33, Number 2 (2005), 583-605.

Dates
First available in Project Euclid: 26 May 2005

Permanent link to this document
https://projecteuclid.org/euclid.aos/1117114330

Digital Object Identifier
doi:10.1214/009053604000001273

Mathematical Reviews number (MathSciNet)
MR2163153

Zentralblatt MATH identifier
1069.62004

Subjects
Primary: 62A01: Foundations and philosophical topics 62F10: Point estimation 62F15: Bayesian inference 62C10: Bayesian problems; characterization of Bayes procedures

Keywords
Estimation invariance parameterization manifold metric Bayesian MAP MMSE mean continuous

Citation

Jermyn, Ian H. Invariant Bayesian estimation on manifolds. Ann. Statist. 33 (2005), no. 2, 583--605. doi:10.1214/009053604000001273. https://projecteuclid.org/euclid.aos/1117114330.


Export citation

References

  • Amari, S. (1985). Differential-Geometrical Methods in Statistics. Lecture Notes in Satist. 28. Springer, Berlin.
  • Barndorff-Nielsen, O. E. (1987). Differential geometry and statistics: Some mathematical aspects. Indian J. Math. 29 335--350.
  • Bernardo, J. M. (1979). Reference posterior distributions for Bayesian inference (with discussion). J. Roy. Statist. Soc. Ser. B 41 113--147.
  • Choquet-Bruhat, Y., DeWitt-Morette, C. and Dillard-Bleick, M. (1977). Analysis, Manifolds and Physics. North-Holland, Amsterdam.
  • Critchley, F., Marriott, P. and Salmon, M. (1994). Preferred point geometry and the local differential geometry of the Kullback--Leibler divergence. Ann. Statist. 22 1587--1602.
  • Efron, B. and Hinkley, D. V. (1978). Assessing the accuracy of the maximum likelihood estimator: Observed versus expected Fisher information (with discussion). Biometrika 65 457--487.
  • Karcher, H. (1977). Riemannian center of mass and mollifier smoothing. Comm. Pure Appl. Math. 30 509--541.
  • Kass, R. E. and Vos, P. W. (1997). Geometrical Foundations of Asymptotic Inference. Wiley, New York.
  • Kass, R. E. and Wasserman, L. (1996). The selection of prior distributions by formal rules. J. Amer. Statist. Assoc. 91 1343--1370.
  • Murray, M. K. and Rice, J. W. (1993). Differential Geometry and Statistics. Chapman and Hall, London.
  • Oliver, J. J. and Baxter, R. A. (1995). MML and Bayesianism: Similarities and differences. Technical Report 206, Dept. Computer Science, Monash Univ.
  • Pennec, X. (1999). Probabilities and statistics on Riemannian manifolds: Basic tools for geometric measurements. In Proc. International Workshop on Nonlinear Signal and Image Processing (NSIP 1999) (A. E. Cetin, L. Akarun, A. Ertuzun, M. N. Gurcan and Y. Yardimci, eds.) 1 194--198. IEEE--EURASIP.
  • Rao, C. R. (1945). Information and accuracy attainable in the estimation of statistical parameters. Bull. Calcutta Math. Soc. 37 81--91.
  • Wallace, C. S. and Boulton, D. M. (1968). An information measure for classification. The Computer Journal 11 185--195.
  • Wallace, C. S. and Boulton, D. M. (1975). An invariant Bayes method for point estimation. Classification Soc. Bull. 3 11--34.
  • Wallace, C. S. and Freeman, P. R. (1987). Estimation and inference by compact coding. J. Roy. Statist. Soc. Ser. B 49 240--265.