Identifiability, Estimability, and Extended Optimality Criteria

  • Luc Pronzato
  • Andrej Pázman
Part of the Lecture Notes in Statistics book series (LNS, volume 212)


Among the major difficulties that one may encounter when estimating parameters in a nonlinear model are the nonuniqueness of the estimator, its instability with respect to small perturbations of the observations, and the presence of local optimizers of the estimation criterion. Classically, those issues are ignored at the design stage: the designs of Chap. 5 are based on asymptotic local properties of the estimator; the approaches of Chap. 6 make use of an assumption (\({\rm H}_{\mathcal{S}}\), p. 183) which allows us to avoid these difficulties. The main message of this chapter is that estimability issues can be taken into account at the design stage, through the definition of suitable design criteria. This forms a difficult area, still under development. Several new notions will be introduced, and a series of examples will illustrate the importance of the geometry of the model.


Design Measure Exact Design Intrinsic Curvature Estimability Function Estimability Condition 
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.


  1. Atkinson, A., K. Chaloner, A. Herzberg, and J. Juritz (1993). Optimal experimental designs for properties of a compartmental model. Biometrics 49, 325–337.zbMATHCrossRefGoogle Scholar
  2. Atkinson, A. and D. Cox (1974). Planning experiments for discriminating between models (with discussion). J. Roy. Statist. Soc. B36, 321–348.Google Scholar
  3. Bates, D. and D. Watts (1980). Relative curvature measures of nonlinearity. J. Roy. Statist. Soc. B42, 1–25.Google Scholar
  4. Chavent, G. (1983). Local stability of the output least square parameter estimation technique. Matematicada Applicada e Computacional 2(1), 3–22.MathSciNetzbMATHGoogle Scholar
  5. Chavent, G. (1987). Identifiability of parameters in the output least square formulation. In E. Walter (Ed.), Identifiability of Parametric Models,  Chapter 6, pp. 67–74. Oxford: Pergamon.Google Scholar
  6. Chavent, G. (1990). A new sufficient condition for the wellposedness of non-linear least-square problems arising in identification and control. In A. Bensoussan and J. Lions (Eds.), Analysis and Optimization of Systems, Lecture Notes in Control and Inform. Sci., vol. 144, pp. 452–463. Springer.Google Scholar
  7. Chavent, G. (1991). New size × curvature conditions for strict quasiconvexity of sets. SIAM J. Control Optim. 29(6), 1348–1372.MathSciNetzbMATHCrossRefGoogle Scholar
  8. Clyde, M. and K. Chaloner (2002). Constrained design strategies for improving normal approximations in nonlinear regression problems. J. Stat. Plann. Inference 104, 175–196.MathSciNetzbMATHCrossRefGoogle Scholar
  9. Demidenko, E. (1989). Optimization and Regression. Moscow: Nauka. [In Russian].zbMATHGoogle Scholar
  10. Demidenko, E. (2000). Is this the least squares estimate? Biometrika 87(2), 437–452.MathSciNetzbMATHCrossRefGoogle Scholar
  11. Hill, P. (1978). A review of experimental design procedures for regression model discrimination. Technometrics 20, 15–21.zbMATHCrossRefGoogle Scholar
  12. Kieffer, M. and E. Walter (1998). Interval analysis for guaranteed nonlinear parameter estimation. In A. Atkinson, L. Pronzato, and H. Wynn (Eds.), MODA5 – Advances in Model-Oriented Data Analysis and Experimental Design, Marseilles (France), June 1998, pp. 115–125. Heidelberg: Physica Verlag.CrossRefGoogle Scholar
  13. Koopmans, T. and O. Reiersøl (1950). The identification of structural characteristics. Ann. Math. Statist. 21(2), 163–181.CrossRefGoogle Scholar
  14. Pázman, A. (1984a). Nonlinear least squares – uniqueness versus ambiguity. Math. Operationsforsch. Statist. Ser. Statist. 15, 323–336.MathSciNetzbMATHGoogle Scholar
  15. Pázman, A. (1993b). Nonlinear Statistical Models. Dordrecht: Kluwer.zbMATHGoogle Scholar
  16. Pázman, A., S. Bilenkaya, J. Bystrický, Z. Janout, and F. Lehar (1969). The problem of un-ambiguity of phase-shift analysis I: Description of a powerful discriminating criterion. Czech J. Physics B19, 882–890.Google Scholar
  17. Pronzato, L. and E. Walter (2001). Eliminating suboptimal local minimizers in nonlinear parameter estimation. Technometrics 43(4), 434–442.MathSciNetCrossRefGoogle Scholar
  18. Ross, G. (1990). Nonlinear estimation. New York: Springer.zbMATHCrossRefGoogle Scholar
  19. Walter, E. (1982). Identifiability of State-Space Models. Berlin: Springer Verlag.zbMATHCrossRefGoogle Scholar
  20. Walter, E. (Ed.) (1987). Identifiability of Parametric Models. Oxford: Pergamon Press.zbMATHGoogle Scholar
  21. Walter, E. and L. Pronzato (1995). Identifiabilities and nonlinearities. In A. Fossard and D. Normand-Cyrot (Eds.), Nonlinear Systems. Modeling and Estimation,  Chapter 3, pp. 111–143. London: Chapman & Hall.Google Scholar
  22. Walter, E. and L. Pronzato (1997). Identification of Parametric Models from Experimental Data. Heidelberg: Springer.zbMATHGoogle Scholar

Copyright information

© Springer Science+Business Media New York 2013

Authors and Affiliations

  • Luc Pronzato
    • 1
  • Andrej Pázman
    • 2
  1. 1.French National Center for Scientific Research (CNRS)University of NiceSophia AntipolisFrance
  2. 2.Department of Applied Mathematics and StatisticsComenius UniversityBratislavaSlovakia

Personalised recommendations