机构地区: 武汉大学空间信息与数字工程研究中心
出 处: 《系统工程》 2004年第9期92-95,共4页
摘 要: 对常用的空间数据挖掘方法进行分析,认为常用的统计学方法对数据的限制较多,且计算量较大,在应用中有一定的局限性。与此相反,神经网络方法由于其固有的自学习能力和抗干扰能力,在空间数所挖掘领域有着广泛的应用前景。最后在空间数据挖掘领域引入GSOM网络,用于空间聚类,通过实例证明,效果较好。 Spatial data mining and knowledge discovery is an integrated technology;it is useful in GIS domains.In the article, the commonly used spatial data mining and knowledge discovery methods have been analyzed. Because most of this method is based on statistical theory, so they have some weakness, on the contract, neural network's inherent ability can overcome this weakness. Finally, the GSOM network has been introduced,it is used in the spatial data mining,and validated by an example.
关 键 词: 空间数据挖掘与知识发现 神经网络 空间聚类
领 域: [自动化与计算机技术] [自动化与计算机技术]