摘要
平均欧氏距离是系统聚类法中使用最普遍的一个距离指标,它是样品分类的重要依据.为了能使分类者更科学地确定分类阈值,对随机平均欧氏距离的统计性质进行了研究,确定了它的概率分布,给出了它的数字特征,为分类阈值的确定提供了理论依据和方法.
Mean Euclidean distance is one of the most common distance index used in hierarchical classification method. It is an important basis for sample classification method. To make categorical threshold measure more scientific, statistical properties of random mean Euclidean distance were researched, its probability distribution was determined, its numerical characteristic was given. This research can provide the theoretical basis and the method for determining categorical threshold measure.
出处
《天津科技大学学报》
CAS
2008年第4期85-88,共4页
Journal of Tianjin University of Science & Technology
基金
天津科技大学引进人才基金资助项目(20030402)
关键词
系统聚类法
随机平均欧氏距离
概率分布
数字特征
分类阈值
hierarchical classification method
random Mean Euclidean distance
probability distribution
numerical characteristic
categorical threshold measure
