カテゴリー変量を被説明変量とする空間影響関数モデル : カテゴリー変量のための空間相関関数と空間影響モデルによる土地利用予測

概要

論文の詳細を見る
We proposed a spatial correlation analysis (SCA) method for analyzing two-dimensional lattice urban data. In the method, however, we can not use categorical variables which are often used in actual urban planning. The aim of this paper is to reconstruct the theory of SCA method for making it possible to use not only continuous variables but also categorical variables. The spatial correlation function (SCF) for measuring the spatial continuity of land use is defined for categorical variables by using dummy variables. One of the three types of the spatial influence function (SIF) models, expressing the spatial relation of urban activities, is defined in the case that the dependent variable is categorical and the independent variables are continuous. The estimated parameters of the model can be obtained by solving an eigenvalue problem. A model of the land use forecasting is described in order to confirm the efficiency of the method.
1992-03-30