Parameter Estimation of Multivariate Geographically Weighted Regression Model Using Matrix Laboratory

摘要

Model of Multivariate Geographically WeightedudRegression (MGWR) is the extension of multivariate spatial liniear model with local observation characters for each observation location. To obtain distribution of the MGWR model, parameter estimation of B(u,v) and variance covariance matrix of error is require to be determined. Besides using mathematical approach, matrixudlaboratory (MATLAB) algorithm can also be used to obtainudparameter estimation of model of MGWR. The MATLAB is audhigh level programming language base on numericaludcomputing technique to solve problems which involvesudmathematical operations with array data bases using matrix and vector formulations. Compared to mathematical approach,MATLAB has some advantages which are extensible and no constraint of variable dimension.