改进的模糊聚类算法

Improved Fuzzy C-means Clustering Algorithm

隐节点中心的选取一直是各种RBF神经网络学习算法面临的主要问题之一,主要针对隐节点中心的选择问题,在研究减法聚类和模糊C-均值聚类算法优缺点的基础上,提出了改进的模糊聚类算法.仿真实验表明,改进的算法增强了网络对离群点的鲁棒性,同时缩短了网络的训练时间.

The selection of RBF hidden node centers is one of the major problems in RBF network learning. Aimed at the difficulty and based on the merits and demerits of the existing RBF learning algorithms such as subtractive clustering and fuzzy C - means clustering, this paper proposes a revised fuzzy clustering which is finally applied to strengthen the robustness of the outliers by the network with the effectiveness proved.