A Study on Artificial Neural Network Based Spatial Data Mining and Knowledge Discovery

摘要

Recent progress in spatial and geographic sciences has led to an explosive growth of spatial data. Extracting relevant knowledge from such volumes of data represents an enormous challenge and opportunity. This paper assesses several approaches to spatial data mining and knowledge discovery, because most of this method is based on statistical theory, so they have some inherent weakness, on the contract, neural networks inherent ability can overcome this weakness. Spatial clustering models were based on the adaptation of a self-adaptive neural network known as Growing Self Organizing Maps(GSOM) .These models provided the basis for the implementation of hierarchical clustering, cluster validity assessment and a method for monitoring learning processes (cluster formation). This framework was tested on a geography Information System (GIS) data set. The results indicate that these techniques may facilitate knowledge discovery tasks by improving key factors.