| 一种聚类算法改进及其在税源分析中的应用 |
| |
| 引用本文： | 郑垂勇,徐利,淦文燕,赵敏.一种聚类算法改进及其在税源分析中的应用[J].解放军理工大学学报,2010,11(2):202-206. |
| |
| 作者姓名： | 郑垂勇 徐利 淦文燕 赵敏 |
| |
| 作者单位： | 郑垂勇,赵敏,ZHENG Chui-yong,ZHAO Min(河海大学,商学院,江苏,南京,210098);徐利,XU Li(河海大学,商学院,江苏,南京,210098;江苏省扬州市国税局,江苏,扬州,225000);淦文燕,GAN Wen-yan(解放军理工大学,指挥自动化学院,江苏,南京,210007) |
| |
| 基金项目： | 江苏省哲学社会科学研究基金资助项目 |
| |
| 摘 要： | 面对大量的涉税数据,为揭示纳税个体间内在的分类特性,发现其中偏离常规的异常纳税行为等,有效提高税源监管力度,针对税源监管的应用需求,分析了数据挖掘中常用聚类算法的特性,提出了基于DENCLUE密度聚类的改进方法。该方法不仅可以发现数据分布的典型聚簇模式,而且可以揭示对税源监管更有价值的离群模式。应用实践验证了该方法的有效性。
|
| 关 键 词： | 数据挖掘 聚类 税源分析 应用 |
Improved cluster ing algorithm and its application in revenue source analysis |
| |
| ZHENG Chui-yong,XU Li,GAN Wen-yan and ZHAO Min.Improved cluster ing algorithm and its application in revenue source analysis[J].Journal of PLA University of Science and Technology(Natural Science Edition),2010,11(2):202-206. |
| |
| Authors: | ZHENG Chui-yong XU Li GAN Wen-yan and ZHAO Min |
| |
| Institution: | Business School,Hohai University,Nanjing 210098,China;Business School,Hohai University,Nanjing 210098,China;Yangzhou Municipal Office of the State Administration of Taxation,Yangzhou 225000,China;Institute of Command Automation,PLA Univ.of Sci.& Tech.,Nanjing 210007,China;Business School,Hohai University,Nanjing 210098,China |
| |
| Abstract: | To ex pose the classif ied features o f tax payers as well as the abnormal ratepaying behaviors and
to enhance the super vision, features of commo n clustering alg orithm in data mining w ere analy zed accor ding
to the applicat ion requirement s. A n improv ed metho d based on DENCLUE density clustering w as proposed.
The new method could uncov er not only the typical clustering mo de of data dist ribut ion but the o utlier
mo de as w el l. T he ex periment in the revenue so urce analy sis show s the validity of this alg orithm |
| |
| Keywords: | data mining clustering revenue source analysis appl icatio n |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《解放军理工大学学报》浏览原始摘要信息 |
| 点击此处可从《解放军理工大学学报》下载免费的PDF全文 |
|
