A discrete step steepest descent search algorithm is developed for the estimation of interregional migration rates in local areas from historical time series. Objective function values are obtained from a constrained least-squares optimization, and it is found that the dimensions of the search space correspond to the number of smoothing operations in the columns of the data input matrix. The iterative algorithm requires the use of symptomatic indicators of intercensal population and is shown to yield improved results over a simple least squares approach.
Get full access to this article
View all access options for this article.
References
1.
ChipmanJ.RaoM., 1964, “The estimation of linear restrictions in regression analysis”, Econometrica, 32, 198.
2.
GoldbergerA., 1964, Econometric Theory (John Wiley, New York).
JudgeG.TakayamaP., 1966, “Inequality restrictions in regression analysis”, Journal of the American Statistical Association, 61, 166.
5.
MeierP., 1969, “Numerische Methoden der Relieferfassung”, Geographica Helvetica, 24, 146–151.
6.
MeierP., 1970, “Stochastic population dynamics for regional water supply and wastewater management decision-making”, Report 13, Water Resources Center, University of Massachusetts, Amherst, Massachusetts.
7.
MeierP., 1972, “Stochastic population projection at design level”, Journal of the Sanitary Engineering Division, American Society of Civil Engineers, 98SA6, 883.
8.
RogersA., 1967a, “Matrix analysis of interregional population growth and distribution”, Papers of the Regional Science Association, 18, 177.
9.
RogersA., 1967b, “Estimating interregional population and migration operators from interregional population distributions”, Demography, 4, 515.
10.
RogersA., 1968, Matrix Analysis of Interregional Population Growth and Distribution (University of California Press, Berkeley).
11.
SiegalJ.HamiltonC., 1952, “Some considerations in the use of the residual method of estimating net migration”, Journal of the American Statistical Association, 47, 475.
12.
TheilH., 1966, “On the use of incomplete prior regression analysis”, Journal of the American Statistical Association, 61, 166.
