Spatial clustering method based on GACUC (greedy agglomerate category utility clustering) and Delaunay triangulation network

Abstract

The invention discloses a spatial clustering method based on a GACUC (greedy agglomerate category utility clustering) and Delaunay triangulation network. Clustering is performed according to spatial attribution and non spatial attribution of spatial data, the maximum similarities of the non spatial attribution exist among spatial elements in each cluster, and the spatial accessibility of the spatial elements is provided. The non spatial attribution clustering is performed by the GACUC (), the non spatial attribution clustering of non numeric type attribute items can be supported, and the application range of the clustering method is expanded; meanwhile, spatial attribution clustering is performed on the basis of the Delaunay triangulation network, the inherent spatial data and non spatial attribution clustering of the spatial data can be implemented, and correspondence and distribution regularities among the spatial elements can be extracted more accurately. The method is simple to implement, automatic computer processing is adopted, data processing and analyzing time is saved, accuracy and availability of the clustering result are improved, and the method has a board application prospect in the field of spatial data extraction.

Description

technical field [0001] The invention belongs to the research field of spatial data mining. In particular, it relates to a spatial clustering method based on GACUC and Delaunay triangulation. Background technique [0002] Spatial clustering method is one of the most active technical methods in the field of spatial data mining research. It has been widely used in many research fields such as geographic information science, epidemiology, biology, and economics, and has produced good social and social benefits. economic benefits. [0003] Spatial clustering is to divide the spatial elements in the spatial database into several clusters with practical significance according to certain division rules, so that the spatial elements in each cluster have the greatest similarity, and the clustering Clusters have the greatest difference. Currently, commonly used clustering methods include partition-based clustering methods, hierarchical-based clustering methods, density-based cluster...

AI Technical Summary

Problems solved by technology

Jiao Limin and other scholars have studied the clustering of spatial elements under the dual constraints of space and attributes, and divided the spatial elements by setting the threshold of attribute distance and spatial distance, but different thresholds will have an important impact on the clustering results, and This method is also unable to evaluate the clustering utility of the clustering results, so the method is not very applicable

[0004] In this context, in the face of the current massive spatial data, the existing spatial clustering methods not only do not have an effective clustering utility evaluation method, but also the application range is constrained by attribute values. It is good to solve the clustering problem of spatial data, and the clustering effect is not good

Images