测绘学报 ›› 2017, Vol. 46 ›› Issue (11): 1910-1918.doi: 10.11947/j.AGCS.2017.20170061

• 地图学与地理信息 • 上一篇    下一篇

基于案例推理的居民地选取方法

谢丽敏1, 钱海忠1, 何海威1, 刘闯2, 段佩祥1   

  1. 1. 信息工程大学地理空间信息学院, 河南 郑州 450052;
    2. 31009部队, 北京 100088
  • 收稿日期:2017-02-09 修回日期:2017-09-01 出版日期:2017-11-20 发布日期:2017-12-05
  • 通讯作者: 钱海忠 E-mail:qianhaizhong2005@163.com
  • 作者简介:谢丽敏(1991-),女,硕士生,研究方向为地图自动综合、空间数据挖掘。E-mail:gis_xlm@163.com
  • 基金资助:

    国家自然科学基金(41571442;41171305)

A Habitation Selection Method by Using Case-based Reasoning

XIE Limin1, QIAN Haizhong1, HE Haiwei1, LIU Chuang2, DUAN Peixiang1   

  1. 1. Institute of Geospatial Information, Information Engineering University, Zhengzhou 450052, China;
    2. 31009 Troop, Beijing 100088, China
  • Received:2017-02-09 Revised:2017-09-01 Online:2017-11-20 Published:2017-12-05
  • Supported by:

    The National Natural Science Foundation of China (Nos. 41571442; 41171305)

摘要:

针对当前中小比例尺地图中居民地选取面临的专家制图经验难以形式化表达的问题,提出一种基于案例推理的居民地选取方法。首先,把制图专家对居民地交互选取结果作为案例对象,挖掘居民地案例的属性特征指标,对属性赋值和归一化处理;然后,采用逐步消元法对居民地最佳属性组合进行选择,并构建源案例库;最后,采用案例推理方法,结合KNN算法,训练案例库确定KNN算法的最佳K值,将新案例与源案例库检索匹配,得出最佳决策结果,进而指导待决策居民地的自动选取。经试验验证,该方法能够较好地还原专家的选取意向,具有较好的抗噪声能力,在面状居民地自动选取中取得了较好的效果。

关键词: KNN算法, 案例推理, 居民地选取, 专家经验

Abstract:

Aiming at the problem that the experience of expert in small and medium scale maps is difficult to be expressed in the habitation selection, this paper puts forward a method based on KNN and case-based reasoning on the habitation selection. First of all, the experts selection result on the habitation as cases. Mining habitation property attribute of cases, attribute assignment and the normalization before construction of source database; then, the method of stepwise elimination is used to select the best attribute combination, and training data to determine the optimal K value of KNN algorithm; finally, combined CBR and KNN algorithm to match the new case with the source case library, and get decision result to guide the automatic selection of habitation. The experimental results show that the proposed method can reduce the selection intention of experts, and has better noise immunity. It achieved a good result in the automatic selection of areal habitation.

Key words: KNN algorithm, case-based reasoning(CBR), habitation selection, expert experience

中图分类号: