Computer Science in Economics and Management

, Volume 4, Issue 4, pp 237–259 | Cite as

Obtaining initial parameter estimates for nonlinear systems using multicriteria associative memories

  • Robert Kalaba
  • Eigh Tesfatsion


Parameter estimation problems for nonlinear systems are typically formulated as nonlinear optimization problems. For such problems, one has the usual difficulty that standard successive approximation schemes require good initial estimates for the parameter vector. This paper develops a simple multicriteria associative memory (MAM) procedure for obtaining useful initial parameter estimates for nonlinear systems. An easily calculated one-parameter family of associative memory matrices is developed; see Equation (25). Each memory matrix is efficient with respect to two criteria: accurate recovery of parameter-output training case associations; and small matrix norm to guard against noise arising from imprecise calculations and observations. For illustration, the MAM procedure is used to obtain initial parameter estimates for a well-known nonlinear economic model, the Solow-Swan growth model. Surprisingly accurate initial parameter estimates are obtained over broad ranges of the family of MAM memory matrices, even when observations are corrupted by i.i.d. or correlated noise.

Key words

Nonlinear estimation artificial neural networks associative memory multicriteria optimization Solow-Swan growth model 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Albert, A., Regression and the Moore-Penrose Pseudoinverse, Academic Press, New York, 1972.Google Scholar
  2. 2.
    Allgower, E. and Georg, K., Simplicial and continuation methods for approximating fixed points and solutions to systems of equations,SIAM Rev. 22 (1980), 28–85.CrossRefGoogle Scholar
  3. 3.
    Bates, D.M. and Watts, D.G.,Nonlinear Regression Analysis and its Applications, Wiley, New York, 1988.Google Scholar
  4. 4.
    Cuyt, A. and Rall, L., Computational implementation of the multivariate Halley method for solving nonlinear systems of equations,ACM Trans. Math. Software 11 (1985), 20–36.CrossRefGoogle Scholar
  5. 5.
    Denison, E.,Trends in American Economic Growth, 1929–1982, The Brookings Institution, Washington, D.C., 1985.Google Scholar
  6. 6.
    Judge, G., Griffiths, W., Hill, R., Lutkepohl, H. and Lee, T.-C.,The Theory and Practice of Econometrics, Wiley, New York, 1985.Google Scholar
  7. 7.
    Kalaba, R. and Tesfatsion, L., Solving nonlinear equations by adaptive homotopy continuation,Appl. Math. Comput. 41 (1991), 99–115.CrossRefGoogle Scholar
  8. 8.
    Kalaba, R., Lichtenstein, Z., Simchony, T. and Tesfatsion, L., Linear and nonlinear associative memories for parameter estimation, MRG Working Paper 8913, Department of Economics, USC, Los Angeles, September 1989, to appear inInform. Sci. Google Scholar
  9. 9.
    Kalaba, R., Kim, M. and Moore II, J., Linear programming and simple associative memories,Appl. Math. Comput. 40 (1990), 203–214.CrossRefMathSciNetGoogle Scholar
  10. 10.
    Kohonen, T.,Self-Organization and Associative Memory, Springer-Verlag, New York, 1988.Google Scholar
  11. 11.
    Kosko, B., Bidirectional associative memories,IEEE Trans. Systems Man Cybernet. SMC-18 (1988), 49–60.CrossRefMathSciNetGoogle Scholar
  12. 12.
    Miller, J., Parameter identification by continuation methods,Appl. Math. Comput. 34 (1989), 17–27.CrossRefGoogle Scholar
  13. 13.
    Murakami, K. and Aibara, T., An improvement on the Moore-Penrose generalized inverse associative memory,IEEE Trans. Systems Man Cybernet. SMC-17 (1987), 699–706.Google Scholar
  14. 14.
    Olivier, P., Optimal noise rejection in linear associative memories,IEEE Trans. Systems Man Cybernet. SMC-18 (1988), 814–815.CrossRefGoogle Scholar
  15. 15.
    Penrose, R., On best approximate solutions of linear matrix equations,Proc. Cambridge Phil. Soc. 52 (1956), 17–20.Google Scholar
  16. 16.
    Poggio, T., On optimal nonlinear associative recall,Biol. Cybernet. 19 (1975), 201–209.CrossRefGoogle Scholar
  17. 17.
    Prescott, E., Robert M. Solow's neoclassical growth model: an influential contribution to economics,Scand. J. Econom. 90 (1988), 7–12.Google Scholar
  18. 18.
    Ramanathan, R.,Introduction to the Theory of Economic Growth, Springer-Verlag, New York, 1982.Google Scholar
  19. 19.
    Rumelhart, D. and McClelland, J. (eds.),Parallel Distributed Processing: Explorations in the Microstructure of Cognition, Vols. 1 and 2, MIT Press, Cambridge, Mass. 1986.Google Scholar
  20. 20.
    Scales, L.,Introduction to Non-Linear Optimization, Springer-Verlag, New York, 1985.Google Scholar
  21. 21.
    Simon, J., Great and almost-great magnitudes in economics,J. Econom. Perspectives 4 (1990), 149–156.Google Scholar
  22. 22.
    Vinod, H. and Ullah, A.,Recent Advances in Regression Methods, Marcel Dekker, New York, 1981.Google Scholar
  23. 23.
    Watson, L., Billups, T. and Morgan, A., G52: HOMPACK: A suite of codes for globally convergent homotopy algorithms,ACM Trans. Math. Software 13 (1987), 281–310.CrossRefMathSciNetGoogle Scholar

Copyright information

© Kluwer Academic Publishers 1991

Authors and Affiliations

  • Robert Kalaba
    • 1
  • Eigh Tesfatsion
    • 2
  1. 1.Departments of Electrical and Biomedical EngineeringUniversity of Southern CaliforniaLos AngelesU.S.A.
  2. 2.Department of Economics and Department of MathematicsIowa State UniversityAmesU.S.A.

Personalised recommendations