置信系数不变下异方差的接受域的优良性及效果研究

The uniqueness and effect on the optimized confidence interval

本文主要采用拉格朗日乘数法,探讨两正态总体方差比的置信区间对应下的假设检验的接受域的优化效果问题。从理论证明和实证的角度得出第一参数和第二参数大小的变化对优化的效果有明显不同的影响,这对实际工作者在实践中调节两参数的大小来得到更为优良的统计结论有较重要的参考意义。

This text mainly analyzed the confidence interval of the ratio of the two normal random variables and obtained the optimized one and proved its existence and uniqueness. The main result was that the first parameter and the second one had significant different effects on the optimized confidence interval, and this result enriched the application of the knowledge and developed a more accurate way.