| 多维统计数据质量检验与异常点识别的模型与方法 |
| |
| 引用本文： | 成邦文,师汉民,王齐庄. 多维统计数据质量检验与异常点识别的模型与方法[J]. 数学的实践与认识, 2003, 33(4): 1-7 |
| |
| 作者姓名： | 成邦文 师汉民 王齐庄 |
| |
| 作者单位： | 华中科技大学管理学院,湖北,武汉,430077 |
| |
| 基金项目： | 国家自然科学基金 ( 79770 0 19)项目 |
| |
| 摘 要： | 本文证明 ,反映社会经济规模大小的多维统计数据近似服从多维对数正态分布 .基于此 ,本文提出了对这类数据的质量及其异常点进行检查和识别的多维对数正态分布检验法 ,并以我国中央部门属研究机构的调查数据为例进行了研究
|
| 关 键 词： | 对数正态分布 统计数据质量检验 异常点识别 |
| 修稿时间： | 2000-11-07 |
The Model and Method for Checking Quality of Multidimensional Statistics and Identifying Outliers from the Data |
| |
| CHENG Bang- wen,SHI Han- min,WANG Qi- zhuang. The Model and Method for Checking Quality of Multidimensional Statistics and Identifying Outliers from the Data[J]. Mathematics in Practice and Theory, 2003, 33(4): 1-7 |
| |
| Authors: | CHENG Bang- wen SHI Han- min WANG Qi- zhuang |
| |
| Abstract: | It is shown that the multidimensional statistical data from survey units,which describe the scale of social and economic system,must be subordinate to multidimensional logarithm normal distribution.Furthermore a method has been given forchecking the quality of the data and identifying the outliers from the data.The paper has also taken the survey data from the research and developmentinstitutions belong to central governmentas an example. |
| |
| Keywords: | logarithm normal distribution quality of multidimensional statistics identifying outliers |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
