摘要
本文采用回归诊断方法来决定自变量的变换和剔除原始数据中的异常点,使原始数据能符合回归模型的假设条件,从而建立起一个模拟“缫丝计价”的最优评茧回归模型。该模型已成功地应用于生产实践。
In the study the method of regrssion diagnosis was used to decide the transformation of the in dependent variable regression equations and to delete the outliers from the initial data in order to make the initial data coincide with the assumed conditions of regression model so that an excellent regression model simulating“Filature Calculated Price”for cocoon evaluation was set up,and the model has being successfully applied in practice.
出处
《浙江农业学报》
CSCD
1990年第2期77-80,共4页
Acta Agriculturae Zhejiangensis
关键词
回归诊断
异常点
评茧模型
蚕茧
Regression diagnosis
Outlier
Model of cocoon evaluation
