摘要
在大量的模糊聚类算法中,模糊C均值聚类算法是应用最为广泛的,然而它存在着一些缺点:对初始值敏感,对噪声数据敏感,容易陷入局部最优。针对以上问题,提出了一种基于粒子群优化的模糊聚类算法,利用粒子群强大的全局寻优能力,这种算法克服了模糊C均值聚类算法的缺点,试验证明,这种算法是一种很有潜力的模糊聚类算法。
Among lots of fuzzy clustering algorithms,the Fuzzy C-means Algorithm (FCM) is the most wide-used. However,FCM has some defects including sensitivity to the initial data,sensitivity to the noise data and getting in the local optimization.In order to overcome those defects,a new PSO-based fuzzy algorithm is put forward.The new PSO- based fuzzy algorithm is evaluated by a data set.Resuh shows that the algorithm has much potential.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第27期150-151,165,共3页
Computer Engineering and Applications
基金
国家自然科学基金资助项目(编号:70373061)