| 多层递阶回归分析 |
| |
| 引用本文： | 李邦宪,张行才.多层递阶回归分析[J].成都信息工程学院学报,1991(2). |
| |
| 作者姓名： | 李邦宪 张行才 |
| |
| 作者单位： | 浙江省金华市气象台
(李邦宪)，浙江省金华市气象台(张行才) |
| |
| 摘 要： | 在经典多层递阶预报模型中,预报因子对模型的贡献大小主要取决于各因子间的相对量值差异,而与预报因子本身的实际预报价值关系甚微。本文针对这一缺陷,提出了一种回归分析与多层递阶方法相结合的统计预报模型——多层递阶回归分析。它既摒弃了经典统计理论中的固定参数预报模型,又较好地消除了由于变量间相对量值差异所造成的贡献差异,充分体现了高相关因子在预报模型中的重要作用。实例计算结果表明,多层递阶回归分析方法的预报效果明显优于经典多层递阶方法和回归分析。
|
| 关 键 词： | 天气学 长期天气预报 多层递阶预报方法 回归分析 |
THE MULTILEVEL RECURSIVE ESTIMATE REGRESSION ANALYSIS |
| |
| Li Bangxian Zhang Hangcai.THE MULTILEVEL RECURSIVE ESTIMATE REGRESSION ANALYSIS[J].Journal of Chengdu University of Information Technology,1991(2). |
| |
| Authors: | Li Bangxian Zhang Hangcai |
| |
| Abstract: | The contribution value of different predictors in the classical multi-level recursive estimate method mainly depends on the relative value contrast, and it shows very little relation with the practical importance of the predictor. In order to avoid the shortcomings of the classical model, a new forecasting model combining the regression analysis with the multi-level recursive estimate regression analysis is presented in this paper. The computating result of an example shows that the forecasting impact of the new model is more stable than that of the classical model. |
| |
| Keywords: | Synoptic meteorology Long-range weather forecast Recursive estimate method Regression |
| 本文献已被 CNKI 等数据库收录！ |
