优化特征加权的FCM算法

Optimization Feature Weighted FCM Algorithm

模糊聚类是一种重要数据分析和建模的无监督方法。在FCM算法中,考虑到样本矢量中各维特征对模式分类的不同影响,本文提出一种优化特征加权的模糊聚类算法,该算法利用主成分分析法提取主要特征向量并根据其对方差的贡献率不同赋予相应权重进行聚类分析。

Fuzzy clustering is a powerful unsupervised method for the analysis of data and construction of models. In the Fuzzy c-Means algorithm, considering the particular contributions of different feature, an optimization feature weighted filthy clustering algorithm is introduced in this paper. By principal component analysis extracting main features and distributing corresponding weight to main features in accordance with their contribution to variance, clustering analysis is implemented by optimization feature weighted fuzzy C clustering algorithm.