| 摘要: |
| 邻域系统是一类具有兼容性与应用性的系统，关于它的粒化一般采用相似度加阈值的方法.基于该方法在阈值确定上的主观性以及不确定计算上结果复杂性，提出一种模糊聚类加统计量的方法，该方法在计算上不仅仍保留Pawlak粗糙集系统一些不确定性度量性质，而且在阈值的确定上更为客观；最后示例说明了该方法的有效性. |
| 关键词: 邻域系统；粗糙集 粒计算；模糊聚类 |
| DOI： |
| 分类号: |
| 基金项目: |
|
| A Granulation Approach and Application in Neighborhood System |
|
LUO Lai peng
|
| Abstract: |
| Neighborhood system is a system with more compatibility and application.Its granulation generally adopts the method of similarity and the threshold.The main shortcomings of the method are the subjectivity of threshold and the complex of calculation results.In this paper, a new approach that adopts fuzzy clustering and statistics is supposed.The approach can not only keep some uncertainty measurement properties of Pawlak rough set system，but also is more objective in the threshold.The example shows the effectiveness of the approach. |
| Key words: neighborhood system rough set granular computing fuzzy clustering |
