摘要
目前,利用智能算法代替贪婪策略求解种子集影响力的方法极大地缩短了运行时间,但一些智能算法因参数设置困难而导致搜索能力不稳定,达不到预期效果,且多数实验仅在小型网络中进行,没有运用到中大型网络.针对此不足,提出一种自适应混沌遗传算法.首先,将一阶容斥激活集引入遗传算法作为适应度函数,目的是在迭代过程中评估种子集的预期影响;然后,利用Logistic混沌序列优化交叉和变异过程中的基因选择;最后,在自适应变异机制下搜索出最优种子集.在4个真实网络上的实验表明,该算法运行效率较高,找到的种子集影响范围较广.
The current method of using intelligent algorithms,which instead of greedy strategy to solve the influence of seed sets,has greatly shortened the running time.However,some intelligent algorithms have unstable search capabilities and fail to achieve the expected results due to difficulty in parameter set,and most of the experiments are only carried out on small networks,not medium and large networks.Aiming at this problem,an adaptive chaotic genetic algorithm is proposed.Firstly,the first-order tolerance activation set is introduced into the genetic algorithm as a fitness function,which the purpose is to evaluate the expected influence of the seed set in the iterative process.Then,the Logistic chaotic sequence is used to optimize gene selection of crossover and mutation.Finally,the self-adaptive mutation mechanism search for the optimal seed set.The experiments on four real networks show that the algorithm runs efficiently and the seed set has a wider range of influence.
作者
张萌
李维华
ZHANG Meng;LI Wei-hua(School of Information Science&Engineering,Yunnan University,Kunming 650500,Yunnan,China)
出处
《云南大学学报（自然科学版）》
CAS
CSCD
北大核心
2021年第2期237-244,共8页
Journal of Yunnan University(Natural Sciences Edition)
基金
国家自然科学基金(61762090)
云南省教育厅科学研究基金(2019J0006).
关键词
自适应混沌遗传算法
影响力最大化
容斥激活集
self-adapting chaos genetic algorithm
influence maximization
inclusion-exclusion activation set
