查询结果:   许爱军.改进ABC算法优化LSSVM的网络流量预测模型[J].计算机应用与软件,2015,32(1):323 - 326.
中文标题
改进ABC算法优化LSSVM的网络流量预测模型
发表栏目
信息技术交流
摘要点击数
1289
英文标题
NETWORK TRAFFIC PREDICTION BASED ON OPTIMISING LSSVM BY IMPROVED ABC
作 者
许爱军 Xu Aijun
作者单位
广州铁路职业技术学院信息工程系 广东 广州 510430     
英文单位
Department of Information Engineering, Guangzhou Institute of Railway Technology, Guangzhou 510430, Guangdong, China     
关键词
网络流量 人工蜂群优化算法 最小二乘支持向量机 预测模型 相空间重构
Keywords
Network traffic Artificial bee colony optimisation algorithm Least square support vector machine Prediction model Phase space reconstruction
基金项目
全国教育信息技术研究“十二五”规划2012年度专项课题(126230653);广东省教育科研“十二五”规划2011年度课题(2011tjk250);中国职业技术教育学会2012-2013年度科研规划项目(203921)
作者资料
许爱军,副教授,主研领域:计算机网络与网络安全。 。
文章摘要
为了提高网络流量预测精度,针对最小二乘支持向量机LSSVM(Least Squares Support Vector Machine)参数优化问题,提出一种改进人工蜂群ABC(artificial bee colony)算法优化LSSVM的网络流量预测模型(ABC-LSSVM)。该模型根据混沌理论对网络流量时间序列进行重构,然后将网络流量预测精度作为优化目标,通过ABC算法找到最优的LSSVM参数,并建立网络流量预测模型,最后采用仿真对比实验测试模型的性能。仿真结果表明,相对于参比模型,ABC-LSSVM解决了LSSVM参数优化的难题,能够更加准确刻画网络流量复杂变化规律,提高了网络流量的预测精度。
Abstract
In order to improve the prediction accuracy of network traffic, we propose a network traffic prediction model (ABC-LSSVM), which is based on optimising the least square support vector machine (LSSVM) with artificial bee colony (ABC) optimisation algorithm and aims at the parameter optimisation issue in regard to least square support vector machine. The model reconstructs the time series of network traffic according to chaos theory, and then takes the network traffic prediction accuracy as the optimisation objective; it finds the optimal LSSVM  parameters by ABC algorithm, and builds prediction model of network traffic. Finally, we use simulative contrasting experiment to test the performance of the model. Simulation results show that compared with other models in the experiment, ABC-LSSVM solves LSSVM parameter  optimisation problem, and can describe the complicated change rules of the network traffic, as well as improves the accuracy of network  traffic prediction.
下载PDF全文