列序分析及其对空气污染的应用

Ordinal Analysis and Its Application to Air Pollution

在数据分析中常用计算相关系数来分析因变量与自变量间的关系,其基本概念是以统计样本来对总体进行估计.列序分析则宁可把问题限制在样本集内,即在已有采样数据的情况下,因变量与m个自变量中哪个自变量联系最密切?哪个又居其次?应该说列序分析更切合实际需要.论证了列序分析的原理、数学表达,同时用大气污染的实测数据进行了计算,对结果进行了分析,且与相关系数的异同作了对比.

In the data analysis,often use correlation coefficient to analyze the relationship between dependent variable and independent variable,the basic concept is to estimate the ensemble statistic according to the statistical sample.However,for an ordinal analysis one prefers to restrict the topic into the sample itself.That is,under the sampling data,which independent variable is tied the first to the dependent variable in the sample？ which is the second？ Should say the ordinal analysis is more suitable to the practice.In this article we argument the principle and the mathematic expression of the ordinal analysis.At the same time,using the observed data of air pollution a real example was calculated,its results were analyzed in comparison with the correlation coefficient,both similarities and differences are demonstrated.