一种改进的可能模糊聚类算法*

通过分析FCM、PCM、IPCM和PFCM等流行的聚类算法和它们在噪声环境下所面临的问题，提出一种概率模糊聚类新算法（SWPFCM），该算法结合样本加权和一种适用于噪音环境下的初始化聚类中心的方法，可以有效地消除噪声对聚类结果的影响。实验表明，SWPFCM算法具有处理大量噪声数据的能力，但对于没有噪声或噪声很少时，效果不明显，当目标样本集中出现噪声时，使用SWPFCM算法聚类将会得到满意的聚类结果。

Improved possibilistic fuzzy clustering algorithm

After analyzing popular clustering algorithms,such as FCM,PCM,IPCM and PFCM,they are sensitive to outliers faults in noisy environments.This paper proposed a new algorithm called sample weighted improved possibilistic fuzzy clustering method(SWPFCM).Based on combination sample weighting and a suitable for noise environment of initialization clustering center method,SWPFCM was less sensitive to outliers.The experimental results with data sets show that SWIPCM algorithm can deal with the amount of noise data,...