| 递进回归-自回归预测方法 |
| |
| 引用本文： | 傅惠民,王治华.递进回归-自回归预测方法[J].机械强度,2006,28(6):874-877. |
| |
| 作者姓名： | 傅惠民 王治华 |
| |
| 作者单位： | 北京航空航天大学,小样本技术研究中心,北京,100083 |
| |
| 摘 要： | 回归分析在工程中有广泛的应用,但是常常由于对问题本身研究的不够全面、深入,或是由于新的影响因素介入等,所选择的自变量不能充分解释因变量,结果导致误差项的均值不为零,此时,采用回归分析进行预测,往往带来较大误差.为此,提出一种递进回归-自回归预测方法,该方法能够充分发挥回归和递进自回归各自的优点,对误差项进行有效补偿.大量计算表明,与传统方法相比,所提方法能显著提高因变量的预测精度.
|
| 关 键 词： | 回归分析 自回归 预测 误差补偿 预测区间 |
| 收稿时间： | 2005-11-09 |
| 修稿时间： | 2005-11-092006-01-15 |
PROGRESSIVE REGRESSION AND AUTOREGRESSION PREDICTION METHOD |
| |
| FU HuiMin,WANG ZhiHua.PROGRESSIVE REGRESSION AND AUTOREGRESSION PREDICTION METHOD[J].Journal of Mechanical Strength,2006,28(6):874-877. |
| |
| Authors: | FU HuiMin WANG ZhiHua |
| |
| Affiliation: | Research Center of Small Sample Technology, Belting University of Aeronautics and Astronautics, Beifing 100083, China |
| |
| Abstract: | Regression analysis has various applications in engineering. The dependent variable cannot be fully explained by chosen independent variables for the flowing two reassons; first, the prablem itself or its physical characteristics cannot be completely researched; second, new influencing factors may intervene. As the trend component of regression residual is not zero, the prediction error may considerably large if regression method is used. A progressive regression and autoregression prediction method is proposed, which can make effective compensation for the trend component of regression residual. Compared with traditional method, this method can significantly improve the prediction precision of the dependent variable. |
| |
| Keywords: | Regression analysis Autoregression Prediction Error compensation Prediction interval |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
