计算机应用与软件2011，Vol.28Issue(2)：34-36,59,4.
基于小波变换和优化的SVM的网络流量预测模型
NETWORK TRAFFIC PREDICTION MODEL BASED ON WAVELET TRANSFORM AND OPTIMISED SUPPORT VECTOR MACHINE
摘要
Abstract
A new network traffic prediction model based on wavelet transform and optimised support vector machine (WsOSVM) is proposed.First, the network traffic is decomposed by non-decimated wavelet transform to acquire the scaling coefficients and wavelet coefficients,and then they are sent individually to different SVM with suitable kernel function for prediction.The parameters of SVM are optimised by adaptive quantum particle swarm optimisation (AQPSO).At last the predictions are combined into the final result by wavelet reconstruction.Experimental results show that the optimised SVM has better generalization performance.The proposed model is suitable for long-term forecast.Compared with the single SVM and the wavelet neural networks model,it has much better prediction precision.关键词
流量预测/(α)Trous/小波变换/SVM/参数优化/量子粒子群引用本文复制引用
周晓蕾,王万良,陈伟杰..基于小波变换和优化的SVM的网络流量预测模型[J].计算机应用与软件,2011,28(2):34-36,59,4.基金项目
国家自然科学基金(60573123). （60573123）
