论文部分内容阅读
判别分析是有力的统计工具,是应用最广的多变量分析方法之一。在土壤学研究中,它也可以应用于数值分类上。通过对土壤初步分类信息进行数学分析,建立判别函数,从而计算被分类客体的判别值或判别归属概率,这样既可以决定未知归类客体的归属,也可以对原来分类的不合理部分进行调整。判别分析要求已知初步分类的一系列土壤客体提供的信息作为统
Discriminant analysis is a powerful statistical tool and is one of the most widely used multivariate analytical methods. In soil science research, it can also be applied to numerical classification. Through the mathematical analysis of the preliminary soil classification information, a discriminant function is established to calculate the discriminant value of the classified object or to determine the ownership probability, which can not only determine the attribution of the unknown classified object, but also adjust the unreasonable part of the original classification. Discriminant analysis requires that the information provided by a series of soil objects of known preliminary classification be used as a system