摘要
将MATLAB模糊工具箱和粗糙集数据处理工具Rosetta结合在一起使用,提出一种基于模糊C聚类划分的并行粗糙集属性值约简算法,将数据集划分到一个个子系统中处理,大大提高了约简的效率。采用聚类算法进行划分,将相似度高的规则放到一个簇中,便于约简,同时由于不同簇的相异程度较高,可以采用直接合并的方式进行全局选择。
Adopt the parallel idea into rough set attribute value reduction, diving the data set into several sub-systems and process-ing at the same time, which greatly improve the reduction efficiency. Using clustering division to put the high similarity rules into acluster to reduce easily, meanwhile, the difference between different clusters contributes to merger directly in global section. Applythe algorithm to corn breeding and it performs better.
出处
《电脑知识与技术》
2015年第4X期176-178,共3页
Computer Knowledge and Technology
基金
国家自然科学基金NSF61272424/F020701
江苏省自然科学基金BK2010277
南通市应用基础研究计划K2010002
关键词
粗糙集
属性值约简
聚类
并行
rough set
attribute value reduction
clustering
parallel
