论文部分内容阅读
求LD_x的最大或然性方法被公认为是最佳方法,但由于计算繁琐而费时往往迫使药理及生物检定工作者要求在不失其精确度的条件下,用简便的方法代替。本文提出用加权近似法求LD_x,并有一套定型的公式以利计算。它略去最大或然性法的叠代过程,使计算大为简化。 该法对最大或然性法是极好的近似,为了说明两者的近似程度,用加权近似法,最大或然性法以及顾汉颐氏提出的简化概率单位法作比较。 最后,文章附录对简化概率单位法的不足提出评论。
The maximum likelihood method of finding LD_x is generally considered the best method, but because of the cumbersome and time-consuming calculation, pharmacists and biometric practitioners are often compelled to require simple and easy methods to replace them without losing their accuracy. This paper proposes to use the weighted approximation to find LD_x, and a set of formulas to facilitate the calculation. It omits the iterative process of the maximum likelihood method, making the calculation greatly simplified. This method is an excellent approximation of the maximum likelihood method. To illustrate the approximation of the two, the method of weighted approximation, maximum likelihood method, and simplified probability unit method proposed by Gu Hanye are compared. Finally, the article appendix comments on the deficiencies of the simplified probability unit method.