Premium
SMALL AREA POPULATION FORECASTING: SOME EXPERIENCE WITH BRITISH MODELS
Author(s) -
OPENSHAW S.,
KNAAP G. A.
Publication year - 1983
Publication title -
tijdschrift voor economische en sociale geografie
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.766
H-Index - 55
eISSN - 1467-9663
pISSN - 0040-747X
DOI - 10.1111/j.1467-9663.1983.tb00976.x
Subject(s) - exponential smoothing , extrapolation , population projection , population , erasmus+ , econometrics , projection (relational algebra) , series (stratigraphy) , projections of population growth , population model , statistics , time series , computer science , geography , operations research , mathematics , population growth , demography , history , paleontology , algorithm , sociology , the renaissance , biology , art history
This study is concerned with the evaluation of the various models including time-series forecasts, extrapolation, and projection procedures, that have been developed to prepare population forecasts for planning purposes. These models are evaluated using data for the Netherlands. "As part of a research project at the Erasmus University, space-time population data has been assembled in a geographically consistent way for the period 1950-1979. These population time series are of sufficient length for the first 20 years to be used to build models and then evaluate the performance of the model for the next 10 years. Some 154 different forecasting models for 832 municipalities have been evaluated. It would appear that the best forecasts are likely to be provided by either a Holt-Winters model, or a ratio-correction model, or a low order exponential-smoothing model."