An application of GIS and Bayesian Network in Studying Spatial Causal Relations between Enterprises and Environmental Factors
The paper intends to employ Geographic Information System (GIS) and Bayesian Network to discover the spatial causality between enterprises and environmental factors in Beijing Metropolis.The census data of Beijing was spatialized by means of GIS in the beginning,and then the training data was made using density mapping technique.Base on the training data,the structure of a Bayesian Network was learnt with the help of Maximum Weight Spanning Tree.Eight direct relations were discussed in the end,of which,the most exciting discovery,Enterprise-Run Society,as the symbol of the former planned economy,was emphasized in the spatial relations between heavy industry and schools.Though the final result is not so creative in economic perspective,it is of significance in technique view due to all discoveries were drawn from data,therefore leading to the realization of the importance of GIS and data mining to economic geography research.
location enterprise GIS Bayesian network economic geography direct relationship causality data mining
SHEN Tiyan LI Xi LI Maiqing
School of Government,Peking University,Beijing,China School of Remote Sensing and Information Engineering,Wuhan University,Wuhan,China
国际会议
第16届国际地理信息科学与技术大会(16th International Conference on GeoInformatics and the Joint Conference)
广州
英文
2008-06-28（万方平台首次上网日期，不代表论文的发表时间）