测绘学报

• 学术论文 • 上一篇    下一篇

粗集方法在地图综合中的应用

李雯静1,邱佳2,龙毅3   

  1. 1. 武汉科技大学资源与环境工程学院
    2. 武汉科技大学
    3. 南京师范大学地理科学学院,虚拟地理环境教育部重点实验室
  • 收稿日期:2010-07-08 修回日期:2010-09-05 出版日期:2012-04-25 发布日期:2012-04-25
  • 通讯作者: 李雯静

Application of Rough Set in Attribute Table without Decision during Map Generalization

  • Received:2010-07-08 Revised:2010-09-05 Online:2012-04-25 Published:2012-04-25

摘要: 将粗集理论与方法应用于地图综合时,根据地图中空间信息的特性及其生成属性决策表的特殊性,提出先利用无监督的离散化方法对各连续型条件属性分别离散化后再根据属性决策表的应用目的对空间目标进行模糊分类添加决策属性的思路;然后利用粗集属性约简的方法计算各条件属性的重要性,并采取动态排序的方法对空间目标的重要性进行判断。该方法可以很好地解决将粗集应用于地图综合中无法预先确定决策属性的问题,对空间点目标的选取实验结果表明该方法是合理有效的。

Abstract: On the basis of the feature of spatial information of map and the particularity of its decision table when apply rough set theory and method to map generalization, an idea of discretize the continuous conditional attributes in the decision table by the method of discretizaton with nonsupervision before adding decision attribute according to the vague categories of spatial objects which obtained by the application purpose of decision table is proposed. Then calculate the importance of the conditional attributes by the method of attribute reduction of rough set to access the importance degree of spatial objects with dynamic sequencing. This process can resolve the problem of can not determine the decision attribute inadvance when apply rough set theory to map generalization, and the experiment result of spatial points selection shows its rationality and validity.