判别分析k最近邻判别分析法2
Discriminant analysis the K-Nearest neighbors discriminant analysis method 2
投稿时间:2025-08-27  修订日期:2025-08-27
DOI:
中文关键词:  k个最近邻  k值选择  距离度量  多数表决规则  判别分析
英文关键词:K-nearest neighbor (k-NN)  K value selection  Distance metrics  Majority voting rule  Discriminant analysis
基金项目:
作者单位地址
胡良平* 军事科学院研究生院 E-mail:lphu927@163.com
摘要点击次数:
全文下载次数:
中文摘要:
      本文目的是介绍与k最近邻判别分析法2有关的基本概念、计算方法、两个实例及其用SAS实现计算的方法。基本概念包括k值选择、距离度量、多数表决规则、懒惰学习、特征空间与归一化;计算方法涉及选择k值、计算距离、排序并选取k个最近邻、统计类别频率和确定多数表决规则;两个实例中的资料分别是“66条康乔水蛇的有关数据”与“七类鱼在6个定量指标上的测定结果”;借助SAS软件,对两个实例中的数据进行了k最近邻判别分析,对SAS输出结果给出了解释。
英文摘要:
      This paper aimed to introduce the fundamental concepts, computational methods, two case studies, and the implementation of k-nearest neighbor (k-NN) discriminant analysis method 2 using SAS. The basic concepts included k value selection, distance metrics, majority voting rules, lazy learning, feature space, and normalization. The computational methods involved k value selection, distance calculation, sorting and selecting the k nearest neighbors, calculating category frequency, and determination of the majority voting rule. The datasets in the two case studies comprised "data on 66 Kangqiao water snakes" and "measurement results of six quantitative indicators from seven fish species." Using SAS software, k-nearest neighbor discriminant analysis was performed on the data from both case studies, and the SAS output results were interpreted.
  查看/发表评论  下载PDF阅读器
关闭