节点文献
基于交叉覆盖设计算法的空间分类挖掘
Spatial Classification Data Mining Based on Alternative Covering Design Algorithm
【摘要】 空间数据库中存储了大量的空间和非空间数据,并且隐含了丰富的知识。为了从空间数据库中发现有价值的模式和知识,文中介绍了空间数据分类挖掘的研究现状和方法,及M-P神经元的超球领域几何意义,然后将多层前向神经网络的交叉覆盖设计算法应用于空间数据分类挖掘中,用来预测和分析森林覆盖类型,实验得到了很好的效果。
【Abstract】 Too much spatial and non-spatial data are stored in spatial database.In order to discover interesting patterns and knowledge from spatial dababase,in this paper,the author introduces present situation and method of the research of spatial classification mining,presents a way of classification analysis and forcasting in spatial data mining using alternative covering design algorithm.The experiment of predicting forest cover types proves that the algorithm is practical.
【关键词】 空间数据挖掘;
分类;
多层前向神经网络;
交叉覆盖设计算法;
【Key words】 spatial data mining; classification; multi-layer forward neural networks; alternative covering design algorithm;
【Key words】 spatial data mining; classification; multi-layer forward neural networks; alternative covering design algorithm;
- 【文献出处】 微机发展 ,Microcomputer Development , 编辑部邮箱 ,2005年04期
- 【分类号】TP311.13
- 【下载频次】112