论文部分内容阅读
本文叙述了如何用相关频数比值作为系数建立二组判别函数的思路及其步骤。强调了数据的转换是判别效果好坏的关键。其原则要求含矿部分的变量值大部分为1,非含矿部分的变量值大部分为0。通过数据的转换可以根据数据的特点发挥人的主观能动性,是本方法的最大优点。采用相关频数比值可以克服小子样带来的麻烦,而且可根据比值的大小直接了解各变量在判别过程中的作用大小。本文并根据南岭花岗岩的化学全分析数据,建立南岭地区含矿与非含矿花岗岩的判别函数作为例子来介绍该方法的应用。
This article describes how to use the relevant frequency ratio as a coefficient to establish two sets of discriminating functions and their steps. Emphasize the data conversion is the key to good or bad. The principle requires that most of the values of the ore-bearing parts are 1 and most of the non-ore-bearing parts are 0. Through the data conversion can play a subjective initiative according to the characteristics of the data, is the greatest advantage of this method. The correlation frequency ratio can overcome the inconvenience caused by the small sample, but also according to the size of the ratio of direct understanding of the role of variables in the process of discrimination. Based on the chemical analysis data of Nanling granite, the discriminant function of ore-bearing and non-ore-bearing granite in Nanling area is established as an example to introduce the application of this method.