摘要
论文在分析聚类与分类方法的基础上,尝试将这两种独立的开采方法集成起来,使两种开采方法能够相得益彰。由此建立的知识发现系统能自动利用已有知识进行分类,同时可较为准确地估计聚类分析所必需的参数,并通过聚类分析来发现和修正错误知识。论文给出了集成方法和实验结果,最后对该集成算法的性能进行了分析。
Clustering and classification are independent methods of data mining.This paper compares relative merits of two methods.And then try to integrate them to construct a effective knowledge mining system.It can classify the unkown by the knowledge it has learned and modify the knowledge database by clustering.This paper also dicusses the experi-ment method,experiment results and analyses the performance of the integrated method.
出处
《计算机工程与应用》
CSCD
北大核心
2003年第10期83-84,共2页
Computer Engineering and Applications
