混合地理加权回归模型算法研究

基金项目: 国家自然科学基金资助项目(49971031);国家教育部长江学者奖励计划资助项目

详细信息

Mining Complete and Correct Frequent Neighboring Class Sets from Spatial Databases

¹ Nanjing Institute of Geography and Limnology,CAS,73 East Beijing Road,Nanjing 210008,China;² Department of Computer Science and Engineering,Harbin Institute of Technology,92 Xidazhi Street,Harbin 150001,China

Funds: 国家自然科学基金资助项目(49971031);国家教育部长江学者奖励计划资助项目

摘要: 以迭代算法为基础,推导出混合地理加权回归模型的常系数(全局参数)和变系数(局域参数)的计算方法,并以上海市住宅小区楼盘销售平均价格为例进行验证。结果表明,混合地理加权回归模型的计算量略大于地理加权回归模型,但对样本数据的拟合更好,局域参数估计更稳健。
- 地理加权回归模型 /
- 混合地理加权回归模型 /
- 空间非平稳性 /
- 迭代算法 /
- 空间分析
Abstract: A recent work has introduced the problem of mining neighboring class sets,where instances of each class of a neighboring class set are grouped using their Euclidean distances from each other.Although the concept of neighboring class sets is a useful one,the effective computation of frequent neighboring class sets is only partially solved.A novel algorithm for mining frequent neighboring class sets from spatial datasets is presented.Compared to the previous algorithm,the algorithm can discover complete and correct frequent neighboring class sets.
- spatial database /
- spatial association rules /
- spatial co-location patterns /
- neighbor-ing class set /
- data mining