Abstract
This article describes EZClimb, a set of SAS/IML steps useful in solving numerical optimization problems. The program uses the method of modified quadratic hill-climbing with either analytical or numerical derivatives to maximize a user-defined criterion function. Modified quadratic hill-climbing is one of the more powerful algorithms known for function optimization but is not widely available outside of the software package GQOPT. The efficacy of the SAS steps is illustrated using Rosenbrock's function, Savin and White's Box-Cox extended autoregressive model, and Klein's Model I.
Similar content being viewed by others
Abbreviations
- BHHH:
-
Berndt, Hall, Hall, and Hausman's optimization algorithm
- FIML:
-
full-information maximum-likelihood
- SAS/IML:
-
SAS Institute's Interactive Matrix Language
- TSP:
-
Time Series Processor software package
References
Amemiya, T., 1983, Non-linear regression methods, in Z. Griliches and M. D. Intriligator (eds), Handbook of Econometrics, Volume 1, North-Holland, New York, pp. 333–389.
Belsley, D. A., 1974, Estimation of systems of simultaneous equations and computational specifications of GREMLIN, Annals of Economic and Social Measurement 3, 551–614.
Belsley, D. A., 1979, On the computational competitiveness of full-information maxium-likelihood and three-stage least-squares in the estimation of nonlinear, simultaneous-equations models, J. Econometrics 9, 315–342.
Belsley, D. A., 1980, On the efficient computation of the nonlinear full-information maximumlikelihood estimator, J. Econometrics 14, 203–225.
Berndt, E. K., Hall, B. H., Hall, R. E., and Hausman, J. A., 1974, Estimation and inference in nonlinear structural models, Annals of Economic and Social Measurement, 3/4, 653–665.
Chow, G. C., 1983, Nonlinear models, chapter 7 in Econometrics, McGraw-Hill, New York.
Chow, G. C., 1968, Two methods of computing full-information maximum likelihood estimates in simultaneous stochastic equations, Internat. Economic Reivew 9, 100–112.
Cramer, J. S., 1986, Econometric Applications of Maximum Likelihood Methods, Cambridge University Press, New York.
Goldfeld, S. M., and Quandt, R. E., 1972, Nonlinear Methods in Econometrics, North-Holland, New York.
Goldfeld, S. M., Quandt, R. E., and Trotter, H. F., 1966, Maximization by quadratic hill-climbing, Econometrrca 34, 541–551.
Goldfeld, S. M., Quandt, R. E., and Trotter, H. F., 1968, Maximization by modified quadratic hill-climbing and other methods, Research Memorandum No. 95, Econometric Research Program, Princeton University.
Greene, W., 1990, An introduction to nonlinear optimization, chapter 12 in Econometric Analysis, Macmillan, New York.
Judge, G. G., Hall, R. C. Griffiths, W. E., Lutkepohl, H., and Lee, T., 1982, Nonlinear regression models, chapter 24 in Introduction to the Theory and Practice of Econometrics, Wiley, New York.
Judge, G. G., Hall, R. C., Griffiths, W. E., Lutkepohl, H., and Lee, T., 1985, Numerical optimization methods, appendix B in The Theory and Practice of Econometrics, 2nd edn, Wiley, New York.
Kmenta, J., 1986, Nonlinear models, in Elements of Econometrics, 2nd edn, Macmillan, New York, section 11-3.
Leyden, D. P., 1990, EZClimb: SAS/IML coding to accompany ‘Modified quadratic hill-climbing with SAS/IML’, Working Paper, Department of Economics, University of North Carolina at Greensboro, Greensboro, North Carolina.
Maddala, G. S., 1977, Econometrics, McGraw-Hill, New York, pp. 171–181.
Moffitt, R., 1986, The econometrics of piecewise-linear constraints: a survey and exposition of the maximum likelihood method, J. Business and Economic Statistics 4, 317–328.
Powell, M. J. D., 1970, A survey of numerical methods for unconstrained optimization, Soc. Industrial and Applied Mathematics Rev. 12, 79–97.
Quandt, R. E., 1983, Computational problems and methods, chapter 12 in Z. Griliches and M. D. Intriligator, (eds), Handbook of Econometrics, Volume 1, North-Holland, New York.
Rosenbrock, H. H., 1961, An automatic method for finding the greatest or least value of a function, Computer J., 3, 175–184.
SAS Institute Inc., 1985, SAS/IML User's Guide, Version 5 Edition, SAS Institute, Inc., Cary, North Carolina.
Savin, N. E., and White, K. J., 1978, Estimation and testing for functional form and autocorrelation, J. Econometrics 8, 1–12.
Spitzer, J. J., 1984, Variance estimates in models with the Box-Cox transformation: Implications for estimation and hypothesis testing, Rev. Economics and Statistics 66, 645–652.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Leyden, D.P. Modified quadratic hill-climbing with SAS/IML. Computer Science in Economics and Management 4, 15–31 (1991). https://doi.org/10.1007/BF00426853
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF00426853