摘要
基于灰色聚类分析理论,采用区间化算子对聚类对象的特性指标规范化,改进了白化权函数的选取方法.通过选择规范的白化权函数,使得每一个评判子类对所有的指标只需设定一种白化权函数.新方法不仅解决了传统的灰色聚类方法计算复杂的缺点,而且有效地克服了当聚类指标的意义、量纲不同,且在数量上悬殊较大时,采用传统的灰色聚类方法可能导致的某些特性指标参与聚类的作用十分微弱的局限性.通过具体的应用实例验证了该方法的有效性.
Based on the principle of grey clustering analysis ,a new grey clustering method is given by adopting interval operator to make the characteristic indexes standardized and improving the method of selecting whitenization weight functions. A new kind of whitenization weight functions is presented to apply in all characteristic indexes. With this new grey clustering method, the difficulty in the calculation is solved and the limitation is effectively overcome that conventional method may lead some characteristic indexes have unreasonable weights when the clustering indexes have different means and dimensions, especially when they are quite different from each other quantitatively. Finally, an illustrative example is given to clarify the method, which shows that the model is feasible, simple and practical.
出处
《西安工程科技学院学报》
2006年第3期369-372,共4页
Journal of Xi an University of Engineering Science and Technology
关键词
灰色系统
灰色聚类
白化权函数
grey system
grey clustering
whitenization weight function
