In this paper we apply the theory of linear associative memories in producing initial parameter estimates for nonlinear iterative approaches. We also propose the use of FEED (Fast and Efficient Evaluation of Derivatives) to evaluate partial derivatives of functions encountered in nonlinear estimation. Suggested methods are presented in the context of calibrating spatial interaction models and are illustrated through numerical examples.
Get full access to this article
View all access options for this article.
References
1.
BattyM, 1976Urban Modeling: Algorithm, Calibration, and Prediction (Cambridge University Press, Cambridge)
2.
BellmanRKalabaR, 1965Dynamic Programming and Modern Control Theory (Academic Press, New York)
FotheringhamAO'KellyM, 1989Spatial Interaction Models: Formulation and Applications (Kluwer, Dordrecht)
5.
FischerM, 1994, “Expert systems and artificial neural networks for spatial analysis and modeling: Essential components for knowledge-based geographical information systems”Geographical Systems1221–235
6.
FischerMGopalS, 1994, “Artificial neural networks: A new approach to modeling interregional telecommunication flows”Journal of Regional Science34502–527
7.
HarrisBWilsonAG, 1978, “Equilibrium values and dynamics of attractiveness terms in production-constrained spatial-interaction models”Environment and Planning A10371–388
KagiwadaHKalabaRRasakhooNSpingarnK, 1986Numerical Derivatives and Nonlinear Analysis (Plenum, New York)
10.
KalabaRKimMMooreJEII, 1991, “Linear programming and associative memories”Applied Mathematics and Computation40203–214
11.
KalabaRKimMMooreJEII, 1992a, “Linear programming and recurrent associative memories”International Journal of General Systems20177–194
12.
KalabaRLichtensteinZSimchnoyTTesfatsionL, 1992b, “Linear and nonlinear associative memories for parameter estimation”Information Sciences6145–66
13.
KalabaRRasakhooN, 1986, “Algorithms for generalized inverses”Journal of Optimization Theory and Applications48427–435
14.
KalabaRTesfatsionL, 1991a, “Obtaining initial parameter estimates for nonlinear systems using multicriteria associative memories”Computer Science in Economics and Management4237–259
15.
KalabaRTesfatsionL, 1991b, “Solving nonlinear equations by adaptive homotropy continuation”Applied Mathematics and Computation4199–115
16.
KalabaRUdwadiaF, 1991, “An adaptive learning approach to the identification of structural and mechanical systems”Computers and Mathematics with Applications2267–75
17.
KohonenT, 1989Self-Organization and Associative Memorythird edition (Springer, New York)
18.
MooreE, 1920, “On the reciprocal of the generalized algebraic matrix”, abstract Bulletin of the American Mathematics Society26394–395
19.
MooreJEIIKimMSeoJ-GWuYKalabaR, 1991, “Linear programming, recurrent associative memories, and feed-forward neural networks”Computers and Mathematics with Applications2271–90
20.
MooreJEIIKimMSeoJ-GKalabaR, 1993, “Obtaining initial parameter estimates for nonlinear systems: Comparing associative memory and neural network approaches”Modelling and Scientific Computing189–115
21.
MooreJEIIKalabaRKimMParkH, 1994a, “Time series and turning point forecasts: A comparison of associative memories and LeSage's econometric techniques”Journal of Regional Science341–26
22.
MooreJEIIKalabaRKimMSeoJ-G, 1994b, “Time series and turning point forecasts: A comparison of artificial neural networks and Bayesian econometric techniques”, copy available from the author
23.
MurakamiKAibaraT, 1987, “An improvement on the Moore—Penrose generalized inverse associative memory”Institute of Electrical and Electronics Engineers, Transactions on Systems, Management and Cybernetics17699–706
24.
OpenshawS, 1992, “Some suggestions concerning the development of artificial intelligent tools for spatial modelling and analysis in GIS”Regional Science2635–51
25.
PenroseR, 1955, “A generalized inverse of matrices”Proceedings of the Cambridge Philosophical Society51406–413
ReillyWJ, 1931The Law of Retail Gravitation (W J Reilly, New York)
28.
ScalesL, 1985Introduction to Non-Linear Optimization (Springer, New York)
29.
SenASmithT, 1995Gravity Models of Spatial Interaction Behavior (Springer, New York)
30.
WilsonAG, 1970aEntropy in Regional Modelling (Pion, London)
31.
WilsonAG, 1970b, “The use of the concept of entropy in system modelling”Operational Research Quarterly21247–265
32.
WilsonAG, 1971, “A family of spatial-interaction models, and associated developments”Environment and Planning A31–32
33.
WilsonAG, 1974Urban and Regional Models in Geography and Planning (John Wiley, New York)
34.
WilsonAGSeniorML, 1974, “Some relationships between entropy maximizing models, mathematical programming models, and their duals”Journal of Regional Science14207–214
35.
XuR, 1995Calibration and Sensitivity Analysis of Nonlinear Systems with Application to Gravity Models PhD Dissertation, School of Urban Planning and Development, Von Kleinsmid Center, University of Southern California, Los Angeles
