摘要
针对空间数据关系复杂的情况,提出了一种改进的多关系数据挖掘结构分类与回归树(S-CART)算法,该算法首先利用空间关联索引表抽取不同主题图层之间的关系原子命题,然后基于逻辑谓词创建多关系二叉决策树,抽取空间关联规则,同时基于我国湖北大冶部分地区土壤污染数据验证算法的有效性。
Regarding the complicated relationship existing in the spatial data, a kind of improved arithmetic based on the S-CART was presented. In this arithmetic, firstly, the relational atoms propositions from different thematic layers were drawn by the Spatial Jion Index Table （SJIT）. And then, based on the predicates, a multi-relational binary decision tree was established and the spatial association rules were drawn out. The efficiency of this arithmetic is validated based on the soil pollution data of some areas of Daye, Hubei.
出处
《计算机应用》
CSCD
北大核心
2008年第3期749-752,共4页
journal of Computer Applications
基金
“十一五”国家科技支撑计划项目(2006BAJ10B01)
