摘要
本文采用模式识别方法推断烧结矿质量。在给出模糊系统聚类分析算法基础上,用软件实现了基于模糊聚类分类器和动态聚类分类器,并用现场实测的样本采用“留一法”分别对这两种分类器性能进行检验。结果表明:模糊聚类分析法对于先验知识较少、样本量不大时,性能较佳。
Application of pattern recognition and AI is a promisive approach to this problem. In the paper, pattern recognition method is used to test the quality of sinter. On the basis of the algorithm of fuzzy clustering analysis, the classifiers based on fuzzy clustering and dynamic state clustering in microcomputer are set up. The performance of the classifiers is tested. The results show that, in the case of lacking priori knowledge, fuzzy clustering analysis is superior to dynamic state clustering analysis when the number of samples is small.
出处
《北京科技大学学报》
EI
CAS
CSCD
北大核心
1993年第5期521-525,共5页
Journal of University of Science and Technology Beijing
基金
国家自然科学基金
关键词
模糊聚类分析
分类器
模式识别
: fuzzy clusteing analysis, classifier, design, properties / pattern recognition, dynamic state clustering