基于改进人工免疫方法的混合模糊聚类算法

摘要

The fuzzy clustering algorithm is sensitive to the initial center and is easily convergence to local optimum. Inspired by the clone selection principle and memory mechanism of the vertebrate immune system, a new hybrid clustering method based on the modified artificial immune theory is presented. It not only adaptively determined the amount and the center’s positions of clustering, but also avoided the local optima and the flaw about sensitive to the initialization. Also, the K-means algorithm is used as a search operator in order to improve the convergence speed. Experimental results indicate the validity and convergence of the proposed algorithm.%为了解决传统模糊聚类算法对初始值敏感、目标函数易陷入局部极小值等问题，将模糊核聚类方法与人工免疫算法相结合，提出了一种基于改进人工免疫方法的混合模糊聚类算法。算法通过借鉴生物免疫系统中的克隆选择原理和记忆机制，自动确定聚类类目及中心位置，同时还集成了模糊c均值搜索算子用于加快收敛速度。仿真实验结果表明了本算法在收敛性、收敛速度和分类性能的有效性。