| 异常观测数据的逐点剔除法 |
| |
| 引用本文： | 王正明,王宝智.异常观测数据的逐点剔除法[J].数学的实践与认识,1997(3). |
| |
| 作者姓名： | 王正明 王宝智 |
| |
| 作者单位： | 国防科技大学数学系，国防科技大学数学系 长沙 410073，长沙 410073 |
| |
| 基金项目： | 国家863项目，湖南省自然科学基金资助 |
| |
| 摘 要： | 本文通过引进一类特殊的统计量和对其统计性质的分析,得到了一种线性回归函数因变量异常观测数据的逐点剔除方法,该方法能很快地认识和剔除含粗大的误差的观测数据,具有低的误判概率。本文还给出了与本文方法配套的数值算法,说明本文方法的数值实例和仿真结果。
|
| 关 键 词： | 线性回归模型 异常数据 特殊统计量 逐点剔除 误判概率 |
One By One-Method of Outliers Rejection In Lineat Regression Model |
| |
| WANG ZHENGMING WANG BAOZHI.One By One-Method of Outliers Rejection In Lineat Regression Model[J].Mathematics in Practice and Theory,1997(3). |
| |
| Authors: | WANG ZHENGMING WANG BAOZHI |
| |
| Abstract: | In this paper, by means of studying the descent degree of the residuls of LS-estimate, we define a statistic to recognize the outliers in observation. Theory analysis shows that this statistic can recognize the polluted observation data with low fals probability. Finally, several famous practical exampals and simulation results are presented to illustrate our method efficiently. |
| |
| Keywords: | LS-estimate outlier false probability robust ectimate |
| 本文献已被 CNKI 等数据库收录！ |
