基于属性权重区间监督的模糊C均值聚类算法

在加权模糊聚类算法中,属性权重确定的合理性是一个重要问题.鉴于用区间数描述决策者推理模糊性的优越性,提出属性权重用区间数表示,由区间层次分析法获得属性对聚类的贡献度,并以该区间为约束条件,提出了可同时获得属性权重和聚类结果的模糊C均值聚类新算法.实验结果表明,该算法以决策者的经验和偏好为监督,可避免迭代计算陷入不必要的局部极小解,能够提高权重分配的合理性,进而得到了更为准确的聚类结果.

A Fuzzy C-Means Algorithm with Interval-Supervised Attribute Weights

The rationality of attribute weights determination is an important issue in the weighted fuzzy clustering algorithms. Because of the advantage of describing the fuzziness of decision maker's inference with interval numbers,attribute weights are represented as intervals,which can be obtained by interval analytic hierarchy process to describe the different contribution of attribute weights for clustering. And a fuzzy C-means algorithm that can obtain attribute weights and clustering results simultaneously is ...