摘要
测定了国产织物样品中的镉含量,对镉含量基础统计学特征进行了描述,采用内核密度分布估计对镉含量数据的多态性进行分析,根据非正态分布的特点,使用Bootstrap法对样本值模拟重复取样,以多次Bootstrap模拟取样的均值与标准偏差作为被考察样品有限单次样本代表值及标准偏差的稳健估计。实践证明:该方法得到的均值与标准偏差作为有限单次样本代表值是合理、有效的,该研究为纺织品禁限用物质的风险评估中禁限用物质含量的表述提供了一种新方法。
Content of cadmium of textile samples collected on domestic market was determined and the content characteristic was investigated by statistical techniques. The polymorphism of data of cadmium content was analyzed using kernel density estimation, and according to the feature of abnormal distribution, bootstrap method was used to simulate re-sampling of the sample. The standing value and standard deviation of cadmium content of these samples obtained by bootstrap re-sampling method served as a sound base for the estimation of the standing value and standard deviation of one of the samples being examined. It was clearly demonstrated by experiments that the above method is rational and effective, providing a new probe to representing the content of banned and restricted substance in risk evaluation of products used in textiles.
出处
《纺织学报》
EI
CAS
CSCD
北大核心
2006年第12期17-20,共4页
Journal of Textile Research
基金
山东出入境检验检疫局重点科研项目(SK05-2004)
关键词
纺织品
镉
代表值
统计学描述
内核密度估计
自举法
textiles
cadmium
standing value
statistical description
kernel density estimation
bootstrap method
