基于改进磷虾群算法的K-means算法

详细信息    查看全文 | 推荐本文 |

英文篇名：K-means Algorithm Based on Improved Krill Herd Algorithm 作者：刘唐 ; 周炜 ; 李志鹏 ; 权文 英文作者：LIU Tang;ZHOU Wei;LI Zhipeng;QUAN Wen;Air Force Engineering University;Xi'an Institute of Finance and Economics,Xingzhi School;Unit 75837 of PLA; 关键词：磷虾群算法 ; 聚类算法 ; 精英引领 ; 最佳聚类数 ; 动态分群 英文关键词：krill herd algorithm;;clustering algorithm;;elitist guiding;;optimal cluster number;;dynamic clustering 中文刊名：XDYX 英文刊名：Journal of Detection & Control 机构：空军工程大学;西安财经学院行知学院;中国人民解放军75837部队; 出版日期：2019-02-26 出版单位：探测与控制学报 年：2019 期：v.41;No.192 基金：国家自然科学基金项目资助(61503407) 语种：中文; 页：XDYX201901015 页数：6 CN：01 ISSN：61-1316/TJ 分类号：78-83

摘要

针对磷虾群算法易陷入局部最优、搜索能力弱及K-means算法易受初始聚类中心选择影响等问题,提出一种基于改进磷虾群算法的K-means算法。该算法通过混沌初始化、动态分群、精英引领和随机变异等策略改进磷虾群算法,并引入最佳聚类数自适应机制,提高了算法的综合寻优能力。实验通过6种基准函数检验了改进磷虾群算法的有效性,用UCI机器学习数据集及人造数据集测试验证了基于改进磷虾群算法的K-means算法的性能。验证结果表明,改进磷虾群算法在保证较快收敛速度的基础上提升了全局寻优能力,与其他算法相比,该算法各方面性能显著提升。
        In order to solve the problem of weak searching ability of standard krill herd algorithm,which is vulnerable to the local optimum, and the K-means algorithm is susceptible to the selection of the initial clustering center, a K-means algorithm based on the improved krill herd algorithm was proposed. The algorithm improved the krill herd algorithm through chaos initialization, dynamic clustering, elite leading and random mutation strategy, and the adaptive mechanism of optimal cluster number was introduced, the comprehensive optimization ability of the algorithm was improved. The validity of the algorithm was verified by six benchmark functions. The performance of the algorithm was validated by UCI machine learning data set and artificial data set test. The verification results showed that the improved krill herd algorithm could improve the global optimization ability on the basis of ensuring the fast convergence speed, and the performance of this algorithm was significantly improved compared with other algorithms.

引文

[1]周志华.机器学习[M].北京:清华大学出版社,2016:197-204.
    [2]杨辉华,王克,李灵巧,等.基于自适应布谷鸟搜索算法的K-means聚类算法及其应用[J].计算机应用,2016,36(8):2066- 2070.
    [3]Gandomi A H, Alavi A H. Krill herd: a new bio-inspired optimization algorithm [J]. Communications in Nonlinear Science and Numerical Simulation, 2012, 17(12): 4831-4845.
    [4]Li Y C, Chih J H,Cheng H Y. Chaotic particle swarm optimization for data clustering [J]. Expert Systems with Applications, 2011,38(21): 14555-14563.
    [5]Satish G, Durga T, Projected Clustering Using Particle Swarm Optimization[J]. Procedia Technology,2012(4): 360- 364.
    [6]Shelokar P S, Jayaraman V K, Kulkarni B D. An ant colony approach for clustering [J]. Analytica Chimica Acta, 2004, 509(2):187-195.
    [7]Chun H C, Angappa G,Kwan H W. An ant colony optimisation-based approach for clustering in a data matrix [J]. International Journal of Operational Research, 2014, 19(4): 407-434.
    [8]耿跃,任军号,吉沛琦. 基于K-means变异算子的混合遗传算法聚类研究[J]. 计算机工程与应用,2011,47(29):151-153.
    [9]曹永春,蔡正琦,邵亚斌.基于K-means的改进人工蜂群聚类算法[J]. 计算机应用,2014,3(1):204- 207.
    [10]于佐军,秦欢.基于改进蜂群算法的K-means算法[J].控制与决策,2018,33(1): 181-185.
    [11]杨燕,靳蕃, Kamel M. 聚类有效性评价综述[J]. 计算机应用研究,2008,25(6):1630-1638.
    [12]Gandomi A H, Siamak T, Faraz T, et al. Krill herd algorithm for optimum design of truss structures [J]. International Journal of Bio-Inspired Computation, 2013, 5(5):281-288.
    [13]黄璇, 郭立红, 李姜,等. 磷虾群算法优化支持向量机的威胁估计[J]. 光学精密工程, 2016, 24(6): 1448-1455.
    [14]牛培峰, 杨潇, 马云鹏,等. 基于改进的磷虾群优化算法的汽轮机初压优化研究[J]. 动力工程学报, 2015, 35(9): 709-713.
    [15]郭伟,高岳林,刘沛. 一种自适应惯性权重的改进磷虾群算法[J].太原理工大学学报, 2016,47(5):651-657.
    [16]李志鹏,李卫忠,杜瑞超.自适应磷虾群优化Elman神经网络的目标威胁评估[J]. 计算机工程与应用,2018,54(7):226-231.
    [17]李丛,胡文军,丁勇,等. 基于改进磷虾群优化的中心极大化KFCM算法在IDS 的应用[J]. 计算机应用研究,2016,33(2):507- 512.
    [18]Xiao H Y, Yun L Z, Wen P Z. A new approach for data clustering using hybrid arti?cial bee colony algorithm [J]. Neurocomputing,2012,97:241-250.