机构地区: 湖南大学电气与信息工程学院
出 处: 《控制与决策》 2009年第12期1881-1884,共4页
摘 要: 不同的模糊分类算法在同一个数据集合上常会产生不同的模糊分类.究竟哪种方法最能揭示数据的真实结构,对此,以模糊分类有效性指标为评价指标,应用层次分析法对各模糊分类进行综合评价,建立了一个模糊分类优选模型.大量实验表明,该优选模型所选出的最优模糊分类,其模式识别率高,能揭示数据的真实结构. Different fuzzy clustering algorithms often generate different fuzzy clusterings over the same data set. Which algorithm can best discover the real structure of the data set is a difficult problem. A selection model for the optimal fuzzy clustering is proposed. This model employs hierarchical analytic process to comprehensively evaluate each alternative fuzzy clustering with multiple cluster validity indexes for fuzzy clusterings and select the optimal one from the alternative fuzzy clusterings. Many experiments show that the optimal fuzzy clustering selected from the alternatives is of the highest pattern recognition rate and perfectly can discover the real structure of the data set.
关 键 词: 模糊分类优选模型 模糊分类 模糊分类有效性指标 层次分析法
领 域: [自动化与计算机技术] [自动化与计算机技术]