首页> 外文期刊>Tropical Animal Health and Production >A comparative study of some estimation methods for parameters and effects of outliers in simple regression model for research on small ruminants
【24h】

A comparative study of some estimation methods for parameters and effects of outliers in simple regression model for research on small ruminants

机译:简单反演模型中反刍动物参数和离群值估计方法的比较研究

获取原文
获取原文并翻译 | 示例
           

摘要

This paper investigated estimation methods: least squares method, M-estimation, Theil method, Least Absolute Deviation method to estimate the parameters of simple regression model in situation that the underlying assumptions of least squares estimation are untenable because of outliers. To compare these methods, the effect of chest girth on body weights of German Farm x Hair crossbred kids at weaning period was examined. Chest girth of kids is independent variable and body weight at weaning period is dependent variable in the model. Mean square error and R(2) value are used to evaluate estimator performance. Because two observation values are outliers and the model estimated from this method have minimum mean square error and maximum R(2) value for different sample sizes (n = 10, 20, 30 or 50), M-estimation method is proposed to predict the parameters of the model.
机译:本文研究了估计方法:最小二乘法,M估计,Theil方法,最小绝对偏差法,以在由于异常值而导致最小二乘估计的基本假设不成立的情况下估计简单回归模型的参数。为了比较这些方法,研究了断奶期胸围对德国农场x杂种孩子体重的影响。在模型中,孩子的胸围是自变量,断奶期的体重是因变量。均方误差和R(2)值用于评估估计器性能。由于两个观测值是离群值,并且对于不同的样本量(n = 10、20、30或50),用此方法估计的模型具有最小均方误差和最大R(2)值,因此建议使用M估计方法来预测模型的参数。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号