基于改进人工蜂群算法的网络突发事件预测  

Prediction of Network Incident Based on Improved Artificial Bee Colony Algorithm

作　　者：蒲国林 刘笃晋 李杰 

机构地区：[1]四川文理学院智能制造学院,达州635000 [2]四川文理学院信息化建设与服务中心,达州635000

出　　处：《计算机与数字工程》2017年第12期2332-2335,2556,共5页Computer & Digital Engineering

基　　金：国家自然科学基金项目(编号:61152003);四川省教育厅重点项目(编号:16ZA0353);四川省教育厅项目(编号:17ZB0377);四川省教育厅项目(编号:16ZB0360);四川文理学院2015年度特色培育一般项目(编号:2015TP001Y)资助

摘　　要：由于网络突发事件预测的复杂性,传统的网络突发事件预测效果并不理想。针对这种情况,对传统人工蜂群算法(ABC)的寻优搜索公式进行了较大改进,将传统的单一搜索公式扩展为雇佣蜂和跟随蜂各一个搜索公式,并根据人工蜂群算法的群体智能涌现原理,将改进的人工蜂群算法(IABC)引入网络群体环境中进行突发事件预测,以此为基础,提出了一种基于改进人工蜂群的网络突发事件预测算法。语料库测评中以K最近邻分类(KNN)、ABC、GABC、IABC三种算法进行实验,实验结果表明,论文所提算法在网络突发事件预测的准确率、召回率、综合性评价指标上都是最优的,因而完全可用于实际。Because of the complexity of the prediction of network emergencies,the prediction results of traditional networkemergencies are not satisfactory. In response to this situation,the traditional search algorithm of artificial bee colony algorithm(ABC)has been improved greatly. The traditional single search formula is extended to a search formula for the employed bees andthe follower bees,respectively. According to the swarm intelligence emergence principle of the artificial bee colony algorithm,theimproved artificial bee colony algorithm(IABC)is introduced into the network group environment for predicting the occurrence ofnetwork emergencies. Based on this,a prediction algorithm for network emergencies based on improved artificial bee colony is pro-posed. The K nearest neighbor classifier(KNN),ABC,GABC、IABC four algorithms are used to evaluate the corpus. The experi-mental results show that the proposed algorithm is optimal in the prediction accuracy,recall rate and comprehensive evaluation in-dex of the network emergency prediction. So it can be used in practice.

关 键 词：网络突发事件 人工蜂群算法 搜索公式 热点事件 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]