基于改进ReliefF算法的特征加权FCM

为改善模糊C均值(FCM)聚类分析算法的性能,减少FCM聚类算法的误分率,提高FCM聚类算法的稳定性,提出了一种改进ReliefF加权FCM(IReliefF-WFCM)聚类算法。IReliefF算法改进了传统ReliefF算法的样本点选择方法,得到了更加稳定有效的特征权值。最后,将该IReliefF-WFCM算法用于数据集等实际数据的聚类分析。结果表明该方法是可行、有效的,为分类模式识别提供了一种误分率小的、稳定的方法。

Characteristic Weighting FCM Based on Improved ReliefF Algorithm

To improve the performance of fuzzy C-mean(FCM) clustering analysis algorithm,reduce the misclassification rate of FCM clustering algorithm and increase the stability of FCM clustering algorithm,this paper presents an improved ReliefF weighting FCM(IReliefF-WFCM) clustering algorithm.IReliefF algorithm improves the sample point selection method of traditional ReliefF algorithm to get more stable and effective characteristic weighting value.At last,this paper uses the IReliefF-WFCM algorithm to perform the clustering analysis of the actual data such as data set.The results indicate that the method is feasible and effective,provides a stable method with small misclassification rate for the classification mode identification.