摘要
针对布谷鸟搜索(CS)算法后期收敛速度慢,传统K-均值算法对初始簇中心选择比较敏感,提出了一种自适应调整的布谷鸟搜索及优化初始K-均值聚类算法(CSSA-OIKM)。首先,由集群度与距离均衡优化选择初始簇中心;其次,融合粒子群算法思想,遵循自适应优化学习策略以均衡CS算法全局与局部精细搜索能力;最后,在改进CS算法的基础上引入自适应度调节步长因子与动态变化发现概率,增强算法收敛性能。通过对经典数据集的仿真实验分析,相比K-均值算法、PSO-K-均值算法及CS-K-均值算法来说,提出的CSSA-OIKM算法能有效提高聚类精确性,且算法稳定性好。
To deal with the problem that the late convergence rate of the cuckoo search（ CS） algorithm was slow and the traditional K-means algorithm was more sensitive to the initial clustering center selection,this paper developed a cuckoo search for self-adaptive adjustment and optimization of initial K-means clustering algorithm（ CSSA-OIKM）. Firstly,it selected the initial cluster center by cluster degree and distance equalization optimization. Secondly,it merged the idea of particle swarm optimization algorithm,followed to the adaptive optimized learning strategy,to equalize the global and local fine search capability of CS algorithm. Finally,based on the improved CS algorithm,it introduced self-adaptive adjust step factor and dynamic change Pa,to enhance the convergence performance of the algorithm. Compared with the K-means algorithm,the PSO-K-means algorithm and the CS-K-means algorithm,the simulation of the classical data set experiment analysis show that the CSSA-OIKM algorithm can improve the accuracy of clustering with better stability.
作者
王日宏
崔兴梅
李永珺
Wang Rihong;Cui Xingmei;Li Yongjun(School of Information ＆ Control Engitnering,Qingdao University of Technology,Qingdao Shandong 266033,China)
出处
《计算机应用研究》
CSCD
北大核心
2018年第12期3593-3597,共5页
Application Research of Computers
基金
国家自然科学基金资助项目(61502262)
山东省研究生教育创新计划资助项目(SDYY16023)
