一种新的带模糊权的粗糙聚类算法

针对粗糙聚类算法缺乏对数据比例变换的鲁棒性的问题，在粗糙聚类的框架下融合模糊聚类的思想，将临界区域中对象的模糊隶属度作为它们对于聚类中心调整的作用权值，得到一种带有模糊权的粗糙聚类算法（fuzzy weighing rough clustering algorithm, FWRCA）．实验表明，该算法不仅对于数据的比例变化具有鲁棒性，且在一定程度上克服了粗糙C均值聚类算法对划分阈值ε的敏感性，在性能上优于传统粗糙C均值聚类算法(如RCMCA)，可应用于水电工程科学等以原型模型为研究手段并有大量需做比例变换的观测数据的领域．

A Novel Rough Clustering Algorithm with Fuzzy Weight

For the problem that the classical rough clustering algorithm is not robust to the scale transformation of datasets,by incorporating the idea of fuzzy clustering under the framework of rough clustering,a fuzzy weighing rough clustering algorithm(FWRCA) is proposed using the fuzzy degree of objects in the boundary region as the weights to adjust the centroids.The experiment shows that the proposed algorithm is not only robust to the scale transformation of datasets,but also is less sensitive to partition threshold than rough C-means clustering algorithm(RCMCA) to a certain extent,therefore,performance of the algorithm is better than that of classic C-means rough clustering algorithms such as RCMCA.It can be applied in those fields(such as hydropower engineering science) in which prototype model is the main research mean and large numbers of observation datasets must be transformed proportionally.