摘要
针对社会网络系统中的社会属性知识没有被充分挖掘,网络结构优化算法学习能力弱的问题,提出了一种Memetic关联学习算法(MRLA)。研究了新算法的基本原理和各个算子,实现了社会属性信息的有效利用。新算法充分结合基于Memetic计算的准确性和基于社会关联学习的快速性,以3个真实社会网络数据集作为测试集,实验结果表明MRLA算法能够有效实现社会网络的聚类分析。
In social networks,the property of society has not been fully exploited.Meanwhile,learning ability for network structure optimization is weak.So a new Memetic Relationship Learning Algorithm(MRLA)has been proposed.This paper studied the fundamentals and basic procedure of MRLA,and effectively utilized the social attribute information.The new algorithm integrated the accuracy of Memetic computation and the quickness of social relational learning.The experimental results of three real-world web data sets show the validity and feasibility of the proposed algorithms.
出处
《复杂系统与复杂性科学》
CSCD
北大核心
2017年第2期89-96,共8页
Complex Systems and Complexity Science
基金
国家重点基础研究发展计划(2013CB329402)
国家自然科学基金(11471004)
中央高校基本科研业务费(GK201603014)
陕西师范大学教学模式创新与实践专项基金(JSJX2016Q014)