摘要
人类认识世界的一种重要方法是将认识对象进行分类,分类可以凭借经验和专业知识来实现,而聚类分析作为一种定量方法.从数据分析的角度,给出了一个更准确、细致的分类工具.作为统计学的一个分支和一种无教师监督的学习方法.聚类分析已有几十年的研究历史,并取得了很多研究成果.目前.聚类学习的典型代表有K-means方法和K-medoid方法等[3].
According to mankind's actual classing procedure for substance,this paper puts forward a new Clustering Learning based on Edge Character (CLEC)approach that is different from some traditional methods. This approach does not depend on the number of class sorts that is very difficult to be gotten accurately and for which the result is rather sensitive. On the other hand, CLEC is fit in the processing to distribution data. The experiments sufficiently proved the efficacy of this approach.
出处
《计算机科学》
CSCD
北大核心
2002年第2期138-139,129,共3页
Computer Science
基金
国家自然科学基金(编号:70071042
60073043)
