摘要
在产品质量检测中的一组试验数据中,有时会出现个别的异常数据,从直观上看,这个数据要比其它数据小许多或者大很多。在处理试验数据时,对于这样的个别异常值,是否要剔除?如果单凭直觉判断,似乎缺乏理论上的依据。本文主要介绍了应用数理统计的方法,对试验数据异常值进行剔除的数理统计方法。
In the process of products quality testing, there are sometimes specific abnormal data in a group of testing data. This data is much bigger or smaller than other data. So, whether should we eliminate the specific abnormal data？ If we only made judgment in intuition, it seems we lacked of theoretical evidence. The article introduced the methods of mathematical statistics with applications and how to eliminate the abnormal value from testing data.
出处
《电子质量》
2007年第2期34-35,共2页
Electronics Quality
关键词
试验
数据处理
数理统计
异常值
剔除
Test
Data processing
Mathematical statistics
Abnormal value
Elimination
