用支持向量机网络实现VBR视频通信量的预测  

VBR Video Traffic Prediction Based on the SVM Networks

作　　者：李素梅[1] 张延炘[1] 常胜江[1] 

机构地区：[1]南开大学信息技术科学学院,天津300071

出　　处：《电子学报》2006年第2期210-213,共4页Acta Electronica Sinica

基　　金：天津市自然科学基金(No.023800811);国家自然科学基金(No.60277022);博士点基金(No.20030055022);国家自然科学基金(No.60477009)

摘　　要：VBR(Variab le B itRate)视频信号具有时变性、非线性和突发性等特点,实现该信号通信量的高精度预测是提高信息传输速度和提高网络带宽资源利用效率的重要手段.针对以上问题,本文提出了一种用于VBR视频通信量预测的差分输入支持向量机(SVM:SupportVectorM achine)网络模型.该网络模型采用结构风险最小化准则,在最小化经验风险的同时,尽量缩小模型预测误差的上界,从而使网络模型具有更好的推广能力.实验结果表明:支持向量机网络模型的预测误差为0.0018,而梯度径向基函数(G rad ient Rad ial Basis Function:GRBF)神经网络模型的预测误差为0.0029.可以看出,支持向量机网络模型的预测精度要比GRBF网络模型的预测精度高出大约40%.A support vector machine neural network is proposed for performing VBR video traffic prediction. It takes differential signal as the input of the network. According to the criteria of structural risk minimization of SVM, the errors between sample - data and model - data are minimized and the upper bound of predicting error of the model is also decreased simultaneously so that the ability of generalization of the model is much improved. The simulation results show that the prediction error of SVM neural networks is 0.0018, while the prediction error of GRBF neural networks is 0.0029. In prediction precision, SVM neural networks model extended 40% than GRBF model.

关 键 词：VBR视频通信量 支持向量机 结构风险 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TP18[自动化与计算机技术—控制科学与工程]