本文主要研究内容
作者陈正鹏（2019）在《基于进化计算的网络社团挖掘研究》一文中研究指出：现实世界中的很多复杂系统,都可以抽象成图的形式,而复杂网络是研究各类系统普适性特征的有效工具。随着网络科学的发展,社团结构被证实为复杂网络的普遍性质。社团挖掘旨在识别网络中特定的社团结构,不仅对分析复杂网络拓扑结构、揭示复杂网络中的规律和预测复杂网络演化具有重要意义,而且有着广泛的应用前景,如社交网络上的信息传播分析,蛋白质功能预测,推荐系统优化等等。但是随着时代的发展,传统社团挖掘算法面临新的考验:1)网络数据规模增加,如何提高静态网络社团挖掘算法的效率是亟需解决的问题;2)网络呈现动态性,如何高效准确地捕捉网络的社团演化过程是亟待解决的问题;3)网络呈现层次性,如何挖掘出复合社团结构从而深层次地了解系统的全貌是目前的研究热点。针对现有的社团挖掘算法面临的静态社团识别效率低、动态社团结构不稳定,复合社团划分不合理的不足,本文围绕进化算法求解网络社团挖掘问题,将社团挖掘转化成优化问题,结合生物仿生模型和多目标优化理论,以“如何提高静态网络社团挖掘算法的效率”、“如何识别高质量且稳定的动态网络社团结构”和“如何识别具有较高全局特征的多层网络复合社团结构”为主线。针对静态网络社团挖掘,提出了一种基于多头绒泡菌仿生模型的离散粒子群优化算法。针对动态网络社团挖掘,提出了一种基于分解的多目标离散粒子群优化算法。针对多层网络社团挖掘,探究了现阶段多目标进化算法在多层网络上的研究现状和应用前景。本文的主要研究贡献可以概括为以下三个方面:(1)针对静态网络社团挖掘问题,采用离散粒子群优化算法求解该问题。设计基于数字串的编码方式利于解码,方便计算适应度值。通过结合网络拓扑邻居的信息和基于贪心的模块度增量策略,修改传统粒子群算法的速度和位置更新公式。并介绍一种多头绒泡菌网络模型(PNM,Physarum network model),可以在一定程度上识别网络社团间和社团内的边。之后,将PNM模型对社团边属性的识别作为先验知识,对粒子群算法的种群初始化进行指导,提出了一种基于多头绒泡菌仿生模型的离散粒子群优化算法。最后,在计算机生成的基准网络和现实世界的网络数据集上进行对比实验,从准确度、鲁棒性和收敛速度方面证明算法的有效性。(2)针对动态网络社团挖掘问题,根据演化聚类框架思想,同时考虑网络社团的结构质量和稳定程度。为了自动确定演化框架中的权重系数和网络的社团数目,同时降低多目标进化算法的计算复杂度,使用基于分解的多目标粒子群优化算法进行问题求解,该算法分别以网络模块化密度和标准化互信息作为优化目标,采用切比雪夫(Tchebycheff)聚合的方式评价个体的适应度值。基于单目标粒子群优化算法,重新设计算法的初始化过程和最优个体状态更新规则,并设计帕累托(Pareto)最优解选取策略,从Pareto最优前端上选取模块化密度最大的解作为当前时刻的社团划分,同时当作下个时刻网络社团挖掘的参照。最后,通过对4类基准网络和7个真实网络数据集上的实验证明了算法的有效性,并对动态社团结构的演变进行了可视化。(3)由于多层网络社团挖掘研究刚刚起步,缺少清晰、统一的定义,本文仅探讨现阶段多层网络社团挖掘的相关理论基础和研究成果,介绍了两种不同类型的多层网络:复合型多层网络和相互依赖型多层网络。针对复合型多层网络的社团挖掘问题进行了形式化定义,汇总现阶段多层网络社团挖掘研究成果,特别是基于多目标进化计算的多层网络社团挖掘算法,为基于进化计算的多层网络社团挖掘研究奠定理论基础。综上,本文在复杂网络理论基础上,结合进化计算和多目标优化理论进行静态和动态网络社团挖掘,并探讨了多目标进化算法在多层网络社团挖掘中的有效性和应用前景。同时,通过大量实验证明了基于进化计算框架的社团挖掘算法的有效性。
Abstract
xian shi shi jie zhong de hen duo fu za ji tong ,dou ke yi chou xiang cheng tu de xing shi ,er fu za wang lao shi yan jiu ge lei ji tong pu kuo xing te zheng de you xiao gong ju 。sui zhao wang lao ke xue de fa zhan ,she tuan jie gou bei zheng shi wei fu za wang lao de pu bian xing zhi 。she tuan wa jue zhi zai shi bie wang lao zhong te ding de she tuan jie gou ,bu jin dui fen xi fu za wang lao ta pu jie gou 、jie shi fu za wang lao zhong de gui lv he yu ce fu za wang lao yan hua ju you chong yao yi yi ,er ju you zhao an fan de ying yong qian jing ,ru she jiao wang lao shang de xin xi chuan bo fen xi ,dan bai zhi gong neng yu ce ,tui jian ji tong you hua deng deng 。dan shi sui zhao shi dai de fa zhan ,chuan tong she tuan wa jue suan fa mian lin xin de kao yan :1)wang lao shu ju gui mo zeng jia ,ru he di gao jing tai wang lao she tuan wa jue suan fa de xiao lv shi ji xu jie jue de wen ti ;2)wang lao cheng xian dong tai xing ,ru he gao xiao zhun que de bu zhuo wang lao de she tuan yan hua guo cheng shi ji dai jie jue de wen ti ;3)wang lao cheng xian ceng ci xing ,ru he wa jue chu fu ge she tuan jie gou cong er shen ceng ci de le jie ji tong de quan mao shi mu qian de yan jiu re dian 。zhen dui xian you de she tuan wa jue suan fa mian lin de jing tai she tuan shi bie xiao lv di 、dong tai she tuan jie gou bu wen ding ,fu ge she tuan hua fen bu ge li de bu zu ,ben wen wei rao jin hua suan fa qiu jie wang lao she tuan wa jue wen ti ,jiang she tuan wa jue zhuai hua cheng you hua wen ti ,jie ge sheng wu fang sheng mo xing he duo mu biao you hua li lun ,yi “ru he di gao jing tai wang lao she tuan wa jue suan fa de xiao lv ”、“ru he shi bie gao zhi liang ju wen ding de dong tai wang lao she tuan jie gou ”he “ru he shi bie ju you jiao gao quan ju te zheng de duo ceng wang lao fu ge she tuan jie gou ”wei zhu xian 。zhen dui jing tai wang lao she tuan wa jue ,di chu le yi chong ji yu duo tou rong pao jun fang sheng mo xing de li san li zi qun you hua suan fa 。zhen dui dong tai wang lao she tuan wa jue ,di chu le yi chong ji yu fen jie de duo mu biao li san li zi qun you hua suan fa 。zhen dui duo ceng wang lao she tuan wa jue ,tan jiu le xian jie duan duo mu biao jin hua suan fa zai duo ceng wang lao shang de yan jiu xian zhuang he ying yong qian jing 。ben wen de zhu yao yan jiu gong suo ke yi gai gua wei yi xia san ge fang mian :(1)zhen dui jing tai wang lao she tuan wa jue wen ti ,cai yong li san li zi qun you hua suan fa qiu jie gai wen ti 。she ji ji yu shu zi chuan de bian ma fang shi li yu jie ma ,fang bian ji suan kuo ying du zhi 。tong guo jie ge wang lao ta pu lin ju de xin xi he ji yu tan xin de mo kuai du zeng liang ce lve ,xiu gai chuan tong li zi qun suan fa de su du he wei zhi geng xin gong shi 。bing jie shao yi chong duo tou rong pao jun wang lao mo xing (PNM,Physarum network model),ke yi zai yi ding cheng du shang shi bie wang lao she tuan jian he she tuan nei de bian 。zhi hou ,jiang PNMmo xing dui she tuan bian shu xing de shi bie zuo wei xian yan zhi shi ,dui li zi qun suan fa de chong qun chu shi hua jin hang zhi dao ,di chu le yi chong ji yu duo tou rong pao jun fang sheng mo xing de li san li zi qun you hua suan fa 。zui hou ,zai ji suan ji sheng cheng de ji zhun wang lao he xian shi shi jie de wang lao shu ju ji shang jin hang dui bi shi yan ,cong zhun que du 、lu bang xing he shou lian su du fang mian zheng ming suan fa de you xiao xing 。(2)zhen dui dong tai wang lao she tuan wa jue wen ti ,gen ju yan hua ju lei kuang jia sai xiang ,tong shi kao lv wang lao she tuan de jie gou zhi liang he wen ding cheng du 。wei le zi dong que ding yan hua kuang jia zhong de quan chong ji shu he wang lao de she tuan shu mu ,tong shi jiang di duo mu biao jin hua suan fa de ji suan fu za du ,shi yong ji yu fen jie de duo mu biao li zi qun you hua suan fa jin hang wen ti qiu jie ,gai suan fa fen bie yi wang lao mo kuai hua mi du he biao zhun hua hu xin xi zuo wei you hua mu biao ,cai yong qie bi xue fu (Tchebycheff)ju ge de fang shi ping jia ge ti de kuo ying du zhi 。ji yu chan mu biao li zi qun you hua suan fa ,chong xin she ji suan fa de chu shi hua guo cheng he zui you ge ti zhuang tai geng xin gui ze ,bing she ji pa lei tuo (Pareto)zui you jie shua qu ce lve ,cong Paretozui you qian duan shang shua qu mo kuai hua mi du zui da de jie zuo wei dang qian shi ke de she tuan hua fen ,tong shi dang zuo xia ge shi ke wang lao she tuan wa jue de can zhao 。zui hou ,tong guo dui 4lei ji zhun wang lao he 7ge zhen shi wang lao shu ju ji shang de shi yan zheng ming le suan fa de you xiao xing ,bing dui dong tai she tuan jie gou de yan bian jin hang le ke shi hua 。(3)you yu duo ceng wang lao she tuan wa jue yan jiu gang gang qi bu ,que shao qing xi 、tong yi de ding yi ,ben wen jin tan tao xian jie duan duo ceng wang lao she tuan wa jue de xiang guan li lun ji chu he yan jiu cheng guo ,jie shao le liang chong bu tong lei xing de duo ceng wang lao :fu ge xing duo ceng wang lao he xiang hu yi lai xing duo ceng wang lao 。zhen dui fu ge xing duo ceng wang lao de she tuan wa jue wen ti jin hang le xing shi hua ding yi ,hui zong xian jie duan duo ceng wang lao she tuan wa jue yan jiu cheng guo ,te bie shi ji yu duo mu biao jin hua ji suan de duo ceng wang lao she tuan wa jue suan fa ,wei ji yu jin hua ji suan de duo ceng wang lao she tuan wa jue yan jiu dian ding li lun ji chu 。zeng shang ,ben wen zai fu za wang lao li lun ji chu shang ,jie ge jin hua ji suan he duo mu biao you hua li lun jin hang jing tai he dong tai wang lao she tuan wa jue ,bing tan tao le duo mu biao jin hua suan fa zai duo ceng wang lao she tuan wa jue zhong de you xiao xing he ying yong qian jing 。tong shi ,tong guo da liang shi yan zheng ming le ji yu jin hua ji suan kuang jia de she tuan wa jue suan fa de you xiao xing 。
论文参考文献
论文详细介绍
论文作者分别是来自西南大学的陈正鹏，发表于刊物西南大学2019-09-24论文，是一篇关于复杂网络论文,社团挖掘论文,动态网络论文,多层网络论文,进化计算论文,多目标优化理论论文,西南大学2019-09-24论文的文章。本文可供学术参考使用，各位学者可以免费参考阅读下载，文章观点不代表本站观点，资料来自西南大学2019-09-24论文网站，若本站收录的文献无意侵犯了您的著作版权，请联系我们删除。
标签：复杂网络论文; 社团挖掘论文; 动态网络论文; 多层网络论文; 进化计算论文; 多目标优化理论论文; 西南大学2019-09-24论文;