| 地理数据线性回归中的稳健估计方法 |
| |
| 引用本文： | 吴健平,张立.地理数据线性回归中的稳健估计方法[J].干旱区地理,1994,17(1):83-88. |
| |
| 作者姓名： | 吴健平 张立 |
| |
| 作者单位： | 华东师范大学地理系!上海200062(吴健平)，上海师范大学地理系!上海200233(张立) |
| |
| 摘 要： | 在线性回归中，常用最小二乘估计求线性方程的回归系数。但最小二乘估计受异常值影响较大，当样本数据存在异常值时，估计出的回归系数会产生较大偏差。稳健估计是最小二乘估计的改进，能在不排除异常数据的情况下，达到减弱异常数据对结果的影响。利用稳健估计提出黄土地区沟谷密度与侵蚀量的回归方程，并和最小二乘估计得到的回归方程比较，前者具有更回归效果。
|
| 关 键 词： | 稳健估计 地理数据 线性回归 |
ROBUST ESTIMATION IN LINEAT REGRESSION OF GEOGRAPHIC DATA |
| |
| Wu Jianping.ROBUST ESTIMATION IN LINEAT REGRESSION OF GEOGRAPHIC DATA[J].Arid Land Geography,1994,17(1):83-88. |
| |
| Authors: | Wu Jianping |
| |
| Abstract: | In linear regression, the least squares estimation is heavily innuenced with outlyers.When there are outlyers in samples,the estimated regressive coefficient will have great errors. Robust estimation is an improvement of least squares estimation,it can reduce innuence of outlyers to result without excluding them. Using robust estimation,the regressive equation or gully density and erosion amount in loess region is obtained,and is compared with that using least squares estimation,it can be found that the former has better regressive result. |
| |
| Keywords: | robust estimation least squares estimation outlyer residual |
| 本文献已被 CNKI 维普 等数据库收录！ |
