计算机科学 ›› 2018, Vol. 45 ›› Issue (5): 201-207.doi: 10.11896/j.issn.1002-137X.2018.05.034
耿焕同,丁洋洋,周利发,韩伟民
GENG Huan-tong, DING Yang-yang, ZHOU Li-fa and HAN Wei-min
摘要： 针对MOEA/D单纯使用邻域更新作为选择策略而造成的个体解的重复更新、缺乏全局适配性等问题,提出了一种兼及全局替换和局部更新策略的新算法,即基于自适应选择策略的改进型MOEA/D(MOEA/D-AS)。算法首先设计了一种新的基于最佳二分图匹配的选择策略(KMS),利用子问题和个体解的匹配关系,从全局角度实现精英个体集的最优选择；然后利用种群的进化信息构造一种匹配紊乱判断机制；最后利用紊乱判断机制,在综合分析邻域更新策略和KMS各自优势的基础上,使算法自适应地选择最合适的选择策略,以提高鲁棒性和优化效率。选取LZ09,DTLZ,CEC09等作为标准测试函数,将改进后的算法MOEA/D-AS与经典MOEA/D系列算法进行对比实验,并以Spread和IGD为性能评估指标。实验结果表明新算法具有更好的收敛性和分布性,验证了自适应选择策略能够有效地指导精英解的选择过程。
| [1] JAIN A K.Data Clustering:50 Years Beyond K-means[C]∥European Conference on Machine Learning and Knowledge Discovery in Databases.Springer-Verlag,2008:651-666. [2] NG A Y,JORDAN M I,WEISS Y.On Spectral Clustering:Analysis and an Algorithm[J]∥Proc Nips,2001,4:849-856. [3] SHI J,MALIK J.Normalized cuts and image segmentation[J].IEEE Trans.Pattern Anal.Mach.Intell,2000,22(8):888-905. [4] BACH F R,JORDAN M I.Learning spectral clustering,with application to speech separation[J].Journal of Machine Lear-ning Research,2006,7(2):2006. [5] OZERTEM U,ERDOGMUS D,JENSSEN R.Mean shift spectral clustering[J].Pattern Recognition,2008,41(6):1924-1938. [6] MA C,WU T,DUAN M Y.Research on clustering algorithm based on degree of membership of K nearest neighbor [J].Computer Engineering and Applications,2016,2(10):55-58.(in Chinese) 马闯,吴涛,段梦雅.基于K近邻隶属度的聚类算法研究[J].计算机工程与应用,2016,52(10):55-58. [7] LU H,FU Z,SHU X.Non-negative and sparse spectral clustering[J].Pattern Recognition,2014,47(1):418-426. [8] YANG X,BAI X,LATECKI L J,et al.Improving Shape Retrieval by Learning Graph Transduction[C]∥European Confe-rence on Computer Vision.Marseille,France,DBLP,2008:788-801. [9] PAVAN M,PELILLO M.Dominant Sets and Pairwise Clustering[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2007,29(1):167-72. [10] LI L,LV J,YI Z.A non-negative representation learning algorithm for selecting neighbors[J].Machine Learning,2016,102(2):133-153. [11] LUXBURG U.A tutorial on spectral clustering[J].Statisticsand Computing,2007,17(4):395-416. [12] CHUANG Y Y.Affinity aggregation for spectral clustering[C]∥IEEE Conference on Computer Vision and Pattern Recognition.IEEE,2012:773-780. [13] ZELNIK-MANOR L.Self-tuning spectral clustering[J].Ad-vances in Neural Information Processing Systems,2004,17:1601-1608. [14] CHANG H,YEUNG D Y.Robust path-based spectral cluste-ring[J].Pattern Recognition,2008,41(1):191-203. [15] ZHANG X,LI J,YU H.Local density adaptive similarity mea-surement for spectral clustering[J].Pattern Recognition Letters,2011,32(2):352-358. [16] ASUNCION A,NEWMAN D J.UCI Machine Learning Repository[EB/OL].http://archive.ics.uci.edu/ml. [17] The MNIST database of handwritten digits.[2015-03-01].http://yann.lecun.com/exdb/mnist. [18] ANANDKUMAR A,FOSTER D P,HSU D,et al.A Spectral Algorithm for Latent Dirichlet Allocation[J].Algorithmica,2015,72(1):193-214. [19] HETTIARACHCHI R,PETERS J F.Multi-manifold LLE lear-ning in pattern recognition[J].Pattern Recognition,2015,48(9):2947-2960. [20] GU L L,PENG L M.Clustering algorithm based on K nearest neighbor of relative density and manifold[J].Computer Science,2016,3(12):213-217.(in Chinese) 古凌岚,彭利民.基于相对密度和流形上k近邻的聚类算法[J].计算机科学,2016,43(12):213-217. |
| No related articles found! |
|
||
首页