论文部分内容阅读
基于生物体免疫和克隆基本原理,提出一种自适应多克隆聚类算法.其核心思想是将多种人工免疫系统算子用于聚类过程,并以亲和度函数为依据自动调整聚类类别.算法引入重组算子来增加抗体种群中个体的多样性以扩大解的搜索范围,避免算法早熟现象.引入非一致变异算子增强局部求解的自适应性、优化局部求解性能,加快算法收敛速度.另外,还利用Markov链证明算法的收敛性.数据仿真实验结果表明该聚类算法能实现合理有效的聚类.
Based on the principles of biological immunity and cloning, an adaptive polyclonal clustering algorithm is proposed, whose core idea is to use a variety of artificial immune system operators in the clustering process and automatically adjust the clustering based on the affinity function The algorithm introduces a recombination operator to increase the diversity of individuals in the antibody population so as to enlarge the searching range of the solution and avoid the premature phenomenon of the algorithm.The introduction of non-uniform mutation operator enhances the self-adaptability of local solution, optimizes the performance of local solution and accelerates the convergence of the algorithm Speed.In addition, we prove the convergence of the algorithm by Markov chain.The simulation results show that the clustering algorithm can achieve reasonable and efficient clustering.