GEOGRAPHICALLY WEIGHTED REGRESSION ON THE ECOLOGICAL FACTORS OF HUMAN LONGEVITY IN GANG-WON PROVINCE, KOREA

摘要

Ordinary least square regression model(OLS) through assumptions the correlations between distribution of longevity population and ecological factors are same in all regions. Therefore it cannot explain the continuous characteristics of ecological data and its spatial variation. Geographically Weighted Regression(GWR) model can quantitatively calculate the spatial similarity of adjacent areas through Geographical Weighting Function. In addition GWR can be described local spatial variation of the distribution of longevity population reflect the environmental characteristics. In this paper we performed a comparative analysis between OLS and GWR model about ecological factors of human longevity that proposed in previous studies.