In this paper, the population projection is estimated by the method of extrapolation which is based on logistic curve. It is required that each time interval of input data is one year. But census is not usually done punctually. So we try to use the ra...
In this paper, the population projection is estimated by the method of extrapolation which is based on logistic curve. It is required that each time interval of input data is one year. But census is not usually done punctually. So we try to use the raw census data and corrected them by appropriate transformation.
For the population projection, we use four kinds of data : Model Ⅰ annual report from 1955 to 1985, Model Ⅱ linear correction of inter census population from 1955 to 1985, Model Ⅲ Compound interest correction on inter census population from 1955 to 1985, Model Ⅳ five years moving average correction of all inter census population. Estimation by model Ⅰ is not good because of large error caused from wrong estimation of population between census.
We get a low projection model
Model Ⅱ P=5,314/(1+1.4904 EXP (-0.050756t)) × 10 exp((4)
from the second try, a high projection model
Model Ⅲ P=5,429/(1+1.5438 EXP (-0.04973t)) × 10 exp(4)
from the third try and a medium projection model
Model Ⅳ P=5,379/(1+1.5190 EXP (-0.050392t)) × 10 exp((4)
from the fourth try.
We can guess at a saturation population and at a demographic transition point by virtue of an estimation of this logistic curve. The saturation population of Korea is estimated as 5,314×10 exp (4) by model Ⅱ in this paper. We have found the fact that the domographic transition point of Korea has been rapidly advancing since 1963.