一种改进森林优化的K-means聚类算法
|
摘要
针对K-means算法易受聚类中心影响而陷入局部最优的问题,提出一种基于改进森林优化算法的K-means聚类算法。首先,将衰减因子引入传统算法中提出一种自适应微量步长方法,以加快算法收敛速度,并改善算法的全局搜索与局部开发能力;然后,结合遗传算法中的算术交叉操作思想,改进传统算法全球播种阶段的选择策略,使得算法能够跳出局部最优,提高算法优化精度。通过基准测试函数实验,验证了改进算法的有效性和优越性。最后,结合改进算法和K-means算法,提出一种新的聚类算法,并通过在UCI数据集上的实验结果表明,提出的聚类算法具有较高的聚类准确率。
In order to solve the problem that the K-means algorithm gets affected by the initial cluster centers easily,this paper proposed a novel K-means clustering algorithm based on improved forest optimization algorithm.Firstly,an adaptive micro-stepping method by taking the attenuation factor into the traditional algorithm was proposed to accelerate the convergence speed and improve the exploration and exploitation capability.Then,the arithmetic crossover in genetic algorithm was introduced to improve selection strategy in the global seeding,which made the algorithm jump out of the local optimum effectively and improved the search precision of the algorithm.The effectiveness and superiority of the improved algorithm were proved with the results on benchmark functions.Finally,combining the improved algorithm with K-means clustering algorithm,a new clustering algorithm was proposed.The results on the UCI dataset show that the improved clustering algorithm has higher clustering accuracy.
In order to solve the problem that the K-means algorithm gets affected by the initial cluster centers easily,this paper proposed a novel K-means clustering algorithm based on improved forest optimization algorithm.Firstly,an adaptive micro-stepping method by taking the attenuation factor into the traditional algorithm was proposed to accelerate the convergence speed and improve the exploration and exploitation capability.Then,the arithmetic crossover in genetic algorithm was introduced to improve selection strategy in the global seeding,which made the algorithm jump out of the local optimum effectively and improved the search precision of the algorithm.The effectiveness and superiority of the improved algorithm were proved with the results on benchmark functions.Finally,combining the improved algorithm with K-means clustering algorithm,a new clustering algorithm was proposed.The results on the UCI dataset show that the improved clustering algorithm has higher clustering accuracy.
引文
[1]PENG K,LEUNG V C M,HUANG Q.Clustering Approach Based on Mini Batch K-means for Intrusion Detection System Over Big Data[J].IEEE Access,2018,6(99):11897-11906.
[2]李云鹏,侯凌燕,王超.基于YOLOv2的复杂场景下车辆目标检测[J].电视技术,2018,42(5):100-106.
[3]周子杰,李勃,董蓉.基于特征提取及聚类算法的增量图片筛选系统[J].电视技术,2017,41(9/10):189-193.
[4]冯仁群山,陈笑蓉.一种基于最大公共子图的文本谱聚类算法[J].贵州大学学报(自然科学版),2018,35(2):82-87.
[5]田诗宵,丁立新,郑金秋.基于密度峰值优化的K-means文本聚类算法[J].计算机工程与设计,2017,38(4):1019-1023.
[6]CHEN T W,IKEDA M.Design and implementation of low-power hardware architecture with single-cycle divider for on-line clustering algorithm[J].IEEE Transactions on Circuits&Systems I Regular Papers,2013,60(8):2165-2176.
[7]周浩理,李太君,肖沙.基于微正则退火算法对K-means聚类算法的优化[J].电视技术,2015,39(17):139-142.
[8]叶志伟,尹宇洁,王明威,等.一种基于杜鹃搜索算法的聚类分析方法[J].微电子学与计算机,2015,32(5):104-110.
[9]贺思云,高建瓴,陈岚.基于改进人工蜂群算法的K-means聚类算法[J].贵州大学学报(自然科学版),2017,34(5):83-87,99.
[10]魏康园,何庆,徐钦帅.基于改进引力搜索算法的K-means聚类[J/OL].计算机应用研究,2019,36(11):1-8.[2018-08-10].http://kns.cnki.net/kcms/detail/51.1196.TP.20180811.1338.094.html.
[11]闫婷,谢红薇.混合细菌觅食和粒子群的K-means聚类算法[J].微电子学与计算机,2016,33(6):59-62,67.
[12]GHAEMI M,FEIZI-DERAKHSHI M R.Forest Optimization Algorithm[J].Expert Systems With Applications,2014,41(15):6676-6687.
[13]CHAGHARI A,FEIZI-DERAKHSHI M R,BALAFAR M A.Fuzzy clustering based on Forest optimization algorithm[J].Journal of King Saud University-Computer and Information Sciences,2018,30:25-32.
[14]GHAEMI M,FEIZI-DERAKHSHI M R.Feature selection using Forest Optimization Algorithm[J].Pattern Recognition,2016,60:121-129.
[15]初蓓,李占山,张梦林,等.基于森林优化特征选择算法的改进研究[J].软件学报,2018,29(9):2547-2558.
[2]李云鹏,侯凌燕,王超.基于YOLOv2的复杂场景下车辆目标检测[J].电视技术,2018,42(5):100-106.
[3]周子杰,李勃,董蓉.基于特征提取及聚类算法的增量图片筛选系统[J].电视技术,2017,41(9/10):189-193.
[4]冯仁群山,陈笑蓉.一种基于最大公共子图的文本谱聚类算法[J].贵州大学学报(自然科学版),2018,35(2):82-87.
[5]田诗宵,丁立新,郑金秋.基于密度峰值优化的K-means文本聚类算法[J].计算机工程与设计,2017,38(4):1019-1023.
[6]CHEN T W,IKEDA M.Design and implementation of low-power hardware architecture with single-cycle divider for on-line clustering algorithm[J].IEEE Transactions on Circuits&Systems I Regular Papers,2013,60(8):2165-2176.
[7]周浩理,李太君,肖沙.基于微正则退火算法对K-means聚类算法的优化[J].电视技术,2015,39(17):139-142.
[8]叶志伟,尹宇洁,王明威,等.一种基于杜鹃搜索算法的聚类分析方法[J].微电子学与计算机,2015,32(5):104-110.
[9]贺思云,高建瓴,陈岚.基于改进人工蜂群算法的K-means聚类算法[J].贵州大学学报(自然科学版),2017,34(5):83-87,99.
[10]魏康园,何庆,徐钦帅.基于改进引力搜索算法的K-means聚类[J/OL].计算机应用研究,2019,36(11):1-8.[2018-08-10].http://kns.cnki.net/kcms/detail/51.1196.TP.20180811.1338.094.html.
[11]闫婷,谢红薇.混合细菌觅食和粒子群的K-means聚类算法[J].微电子学与计算机,2016,33(6):59-62,67.
[12]GHAEMI M,FEIZI-DERAKHSHI M R.Forest Optimization Algorithm[J].Expert Systems With Applications,2014,41(15):6676-6687.
[13]CHAGHARI A,FEIZI-DERAKHSHI M R,BALAFAR M A.Fuzzy clustering based on Forest optimization algorithm[J].Journal of King Saud University-Computer and Information Sciences,2018,30:25-32.
[14]GHAEMI M,FEIZI-DERAKHSHI M R.Feature selection using Forest Optimization Algorithm[J].Pattern Recognition,2016,60:121-129.
[15]初蓓,李占山,张梦林,等.基于森林优化特征选择算法的改进研究[J].软件学报,2018,29(9):2547-2558.