Abstract
Two measures of the influence of the prior distribution p(θ) in Bayes estimation are proposed. Both involve comparing with alternative prior distributions proportional to p(θ)s, for s ≥ 0. The first one, the influence curve for the prior distribution, is simply the curve of parameter values which are obtained as estimates when the estimation is made using p(θ)s instead of p(θ). It measures the overall influence of the prior. The second one, it the influence rate for the prior, is the derivative of this curve at s = 1, and quantifies the sensitivity to small changes or inaccuracies in the prior distribution. We give a simple formula for the influence rate in marginal posterior mean estimation, and discuss how the influence measures may be computed and used in image processing with Markov random field priors. The results are applied to an image reconstruction problem from visual field testing and to a stylized image analysis problem.
Similar content being viewed by others
References
Bengtsson B, Heijl A, Olsson J, Rootzén H (1997) A new generation of algorithms for computerized threshold permetry, SITA. Acta Ophthalmol Scand 75:368–375
Bengtsson B, Heijl A, Olsson J (1998) Evaluation of a new threshold visual field strategy, SITA, in normal subjects. Acta Ophthalmol Scand 76:165–169
Berger JO (1980) Statistical decision theory, foundations, concepts, and methods. Springer, Berlin Heidelberg New York
Berger JO (1990) Robust Bayesian analysis: sensitivity to the prior. J Stat Plan Inference 25:303–328
Bose S (1994) Bayesian robustness with more than one class of contaminations. J Stat Plan Inference 40:177–187
Carlin BP, Chaloner KM, Louis TA, Rhame FS (1995) Elicitation, monitoring, and analysis for an AIDS clinical trial, Lecture Notes in Statistics 105, 48–89, Springer, Berlin Heidelberg New York
Clarke B, Gustafson P (1998) On the overall sensitivity of the posterior distribution to its input. J Stat Plan Inference 71:137–150
Delampady M, Dey DK (1994) Bayesian robustness for multiparameter problems. J Stat Plan Inference 40:375–382
Insua RI, Ruggeri F ed. (2000) Robust Bayesian analysis. Lecture Notes in Statistics, 152, Springer, Berlin Heidelberg New York
Kindermann R, Snell JL (1980) Markov random fields and their applications. American Mathematical Society, Providence
Lehmann EL (1959). Testing statistical hypothesis. Wiley, New York
O’Hagan A (1994) Kendall’s advanced theory of statistics. Vol. 2B, Bayesian inference. University Press, Cambridge
Olsson J, Rootzén H (1994) An image model for quantal response analysis in perimetry. Scand J Stat 21:375–387
Ruggeri F, Wasserman L (1993) Infinitesimal sensitivity of posterior distributions. Can J Stat 21:195–203
Winkler G (1995) Image analysis, random fields and dynamic Monte Carlo methods. Springer, Berlin Heidelberg New York
Acknowledgment
We would like to thank anonymous referees and an associate editor for useful advice which improved the paper considerably.
Author information
Authors and Affiliations
Corresponding author
Additional information
Research supported by the Bank of Sweden Tercentenary Foundation.
Rights and permissions
About this article
Cite this article
Rootzén, H., Olsson, J. On the influence of the prior distribution in image reconstruction. Computational Statistics 21, 431–444 (2006). https://doi.org/10.1007/s00180-006-0004-1
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00180-006-0004-1