Gis-based Spatial Clustering Algorithm, The Research And Application

Introducing the concept, pattern and method of DM and KDD into the field of geograhpy data analysis and finding new method that is good for the data mining of the field of geography, We can combine these with GIS and use the storage and management function of GIS to make the limited data in GIS bacome the limitless knowledge. For doing so, we can effectively handle the huge data, advance the automatization and intelligence level of the geography data analysis and offer powerful analysis tools for the globle change and sustainable development.The spatial clustering analysis is a very important method in the SDM. By introducing the method of spatial clustering analysis into the spatial database of GIS, we can plot the spatial object of GIS by certain rules and find the proper cluster. Then, we can analyze all the attribute character of every cluster.There are many clustering arithmetic at present. Every method has its character and strongpoint. With the fast development and widely application of GIS, how to draw out useful spatial relation and knowledge from GIS is of great significance.However, the arithmetic and application concerning the GIS are not so many, it is under research. This thesis takes the character of data and its application in the soil programing of GIS as background, begins with the spatial relation of the data, and concentrates on how to realize the available clustering of the adjoining spatial terra. Surrounding this question, this thesis mainly do some task as follows.(1) Knowing about the GIS(2) Discussing the evolvement and development of DM and SDM.(3) Introducing and comparing some kinds of classical clustering arithmetic.(4) On the basis of GIS, we bring forward a new clustering arithmetic and realize the computer simulation experiment.(5) Improving the arithmetic from the single condition to the multi-condition. Extending the application of the arithmetic.