基于动态加权LS-SVM的网络流量混沌预测

网络流量是具有复杂非线性、不确定时变性的混沌时间序列.为提高标准最小二乘支持向量机的预测精度与自适应性,提出一种基于动态加权最小二乘支持向量机的网络流量混沌预测方法.该方法在标准LS-SVM回归机的训练样本误差设置时间权,增强对非线性样本的逼近能力.然后结合滚动窗与迭代求逆法实现模型动态在线校正,进而克服网络变化时的累积误差.仿真实验结果表明,相对常规LS-SVM,该模型能降低预测误差、减少计算时间,实现高精度实时混沌流量估计.

Chaotic Prediction for Network Traffic Based on Dynamic Weighted Least Squares Support Vector Maehine

Network traffic is a chaotic time series data of complex nonlinear, indefinitely time-varying. In order to improve the accuracy and adaptation of standard least square support vector machine, a network traffic prediction model is proposed based on dynamic weighted least squares support vector machine. In this model, the time weighting factors are set up in standard LS-SVM, enhancing approximation ability of nonlinear sample, and suppressing singularity of time series. Then, the data in sliding time window, iterative solution is used to updates the traffic model in terms of cumulative error by the change of network. The simulation results show that, compared with traditional algorithm, the proposed model is able to decrease error of forecasting and realize the real-time estimation of network traffic.