摘要
在传统三角白化权函数的基础上,构建一个新的正弦曲线形式的白化权函数,以提高聚类对象划分为其所属灰类的聚类系数,继而建立一种灰色聚类评价改进方法.经仿真实验发现,改进方法能够有效降低聚类信息熵的值,并能提高聚类对象的归属性.通过引用研究生招生实际数据,分析验证了改进的灰色聚类评价方法在招生质量评价应用中的可行性和有效性.
On basis of the traditional triangular whitenization weight function, a new form of whitenization weight function based on sine curve is built in this paper. It can improve the clustering ratio of clustering object belonged to gray class. Then an improved method of gray clustering method is given. Simulation experimental results show that the improved method can effectively reduce the value of information entropy and increase the certainty of clustering objects attribution. With the actual data of graduate student＇s enrollment cited, the lea sibility and effectiveness of the improved gray clustering evaluation method is detailedly analyzed.
出处
《延边大学学报（自然科学版）》
CAS
2015年第4期318-325,共8页
Journal of Yanbian University（Natural Science Edition）
基金
福建省自然科学基金资助项目(2011J01357)
宁德师范学院服务海西资助项目(2012H405)
福建省大学生创新创业训练计划项目(201410398059)
关键词
正弦曲线
白化权函数
灰色聚类
招生质量评价
sine curve
whitenization weight function
gray cluster
enrollment quality evaluation
