摘要
针对当前网络流量非线性时变、混沌等特点以及现有的基于支持向量机(support vector machine,SVM)网络流量预测模型存在预测稳定性不好、精度较低等问题,采用模糊层次分析法对SVM预测模型进行改进,首先使用模糊层次分析法对SVM的σ和C参数进行寻优,然后用寻找到的最优参数来训练SVM,最后建立预测模型,预测网络流量.实验结果表明,本文方法不但可以较好的跟踪网络流量变化趋势,从而可以使网络流量的预测值与实际非常接近,而且预测误差变化范围波动小,是一有效的并且预测精度高的网络流量预测方法.
For the current network traffic has characteristics of nonlinear time-varying,chaos and also has such problems as predict the stability is bad and the low accuracy problem existing many SVM network traffic prediction model, SVM prediction model was im- proved in this paper, using fuzzy hierarchy analysis to SVM for parameter optimization, and first to find the optimal parameters and of SVM training, then establish a prediction model and predict the network traffic. The experimental results show that the method can bet- ter tracking network traffic trends,so that the predict value is close to actual value of the network traffic, and the prediction error range fluctuation is small, and the method in this paper is a high forecast precision and effective network traffic prediction method.
出处
《小型微型计算机系统》
CSCD
北大核心
2015年第6期1261-1264,共4页
Journal of Chinese Computer Systems
基金
河南省科技计划重点项目(102102210416)资助
作者简介
作者简介:王启明。男,1980年生,硕士,讲师,研究方向为软件工程、算法、物联网;
单冬红,女,1976年生,硕士,副教授,研究方向为数据挖掘、网络管理;
赵伟艇,男,1966年生,硕士,教授,研究方向为网络安全、网络管理.