这项研究是为了估计未知的总体方差而完成的,用于均值和方差的变量。为了完成这项任务,利用辅助变量的已知描述提出了一类新的广义的鲁棒方差估计器,例如,中档,Hodges-LehmannMean,三均值,十分位数的意思是,偏度系数,四分位数间距,第一个四分位数,峰度系数,半四分位数平均,分位数范围和平均值,等。这些辅助变量的常规度量提高了在无替换的简单随机抽样(SRSWOR)方案下建议类的准确性。诸如偏置之类的属性,均方误差(MSE),和建议类的最小MSE导出到一阶近似。还从理论上得出了发达估计器类相对于现有估计器的优越性条件。最后,文章背后的动机也完成了数字表示。通常的方差估计器被认为是比较数字插图中所有考虑的估计器的基准。结果表明,建议的类比通常的方差估计器和所有其他有关现有估计器的想法更好。
This study is completed for the estimation of unknown population variance for the variable of mean and variance of interest. To accomplish this task, a new generalized class of robust kind of variance estimators proposed utilizing known descriptives of auxiliary variable, for example, Mid-range, Hodges-Lehmann Mean, Tri-mean, deciles mean, coefficient of skewness, interquartile range, first quartile, coefficient of kurtosis, semi-interquartile average, inter decile range and Mean, etc. These conventional measures of auxiliary variable improve the accuracy of the suggested class under simple random sampling without replacement (SRSWOR) scheme. The properties such as the bias, mean square errors (MSE), and least MSE of the suggested class are derived up to first order of approximation. The superiority conditions of the developed class of estimators over existing estimators are also made out theoretically. Finally, numerical representation is also completed for the motivations behind the article. The usual variance estimator is considered as a benchmark for comparing all considered estimators in numerical illustration. The results have been indicated that the suggested class is performing better than the usual variance estimator and all other thoughts about existing estimators.