关于漏失数据弥补的几点意见  

A Few Ideas on Redeeming of Leak Data

作　　者：黄必恒[1] 

机构地区：[1]浙江林学院基础部,临安311300

出　　处：《大学数学》2004年第1期109-113,共5页College Mathematics

摘　　要：漏失数据的弥补,通常采用使误差项的平方和达到最小的原则来确定,这样弥补的数据,有时会出现负值等异常情况.本文提出采用相邻小区的平均值来代替漏失数据,或用回归分析的方法分析试验结果,同样能得到最小二乘法相一致的结论,但不会出现负值.Redeeming of leak data is usually confirmed by using the principle of making the square sum of error terms to reach least. But he redeeming data by this method sometimes appear some abnormal phenomena such as negative values. In this paper, by using the average value of border small sections to replace the leak data or using the method of regression analysis to analysis the experimental results, we can obtain the accordant conclusion that is obtained by using method of minimum square. Further more using the two methods will not appear negative values.

关 键 词：漏失数据 最小二乘法 均值 回归分析 

分 类 号：O212.1[理学—概率论与数理统计]