摘要
鉴于模糊C-均值类型算法(FCM算法)对初始中心敏感的问题,提出了一种基于遗传算法和模糊聚类的文本分类方法。采用遗传算法初始聚类中心,并在适应度的计算中采用了一个可变值,用户可以在文本直接聚类时更改该值,产生用户满意的属性约简结果,极大地提高了系统的分类精度。最后通过实验给出了该算法性能的测试结果。
In view of the nature of fuzzy C- means algorithm （FCM） is sensitive to initial value,presented a method of text categorization based on genetic algorithm and fuzzy clustering, using genetic algorithms to initial cluster centres. And variable values was adapted in the fitness computation procedure, users could change the value of cluster in the text in order to obtain the customer satisfacted properties, thus greatly improve the fusion system＇s classification accuracy. The experimental results of the algorithm are given at the end of the paper through experiments.
出处
《计算机技术与发展》
2009年第4期131-133,142,共4页
Computer Technology and Development
基金
国家自然科学基金重点项目(60736014)
关键词
模糊聚类
遗传算法
文本分类
FCM
fuzzy clustering
genetic algorithm
text categorization
FCM
