基于自适应遗传算法LS-SVM的网络流量预测

PREDICTING NETWORK TRAFFIC BASED ON ADAPTIVE GENETIC ALGORITHM AND LS-SVM

网络流量受众多因素的影响并且具有复杂的非线性特点,因此网络流量的预测和分析是一个很复杂的问题,最小二乘支持向量机能够成功地解决非线性问题并应用于网络流量的预测和分析。提出一种最小二乘支持向量机模型,将自适应遗传算法用于最小二乘支持向量机参数寻优,并将该模型用于网络流量的预测和分析。对比实验表明,基于最小二乘支持向量机的网络预测模型具有更强的预测能力,在网络流量预测中有一定的实用价值。经实例验证,该模型预测精度高。

Network traffic is affected by many factors and has complex nonlinear characteristic,therefore to predict and analyse network traffic is a very complicated problem,but the least squares support vector machine（LS-SVM） can successfully solve the nonlinear issues and be applied to network traffic foretell and analysis.In this article we proposed a model of least squares support vector machine,which applies the adaptive genetic algorithm in parameter search optimization of the LS-SVM,and used this model to predict and the analyse network traffic. Comparative experiments show that the network prediction model based on least squares support vector machine has stronger predictive ability, and has certain applied value to the network traffic prediction.As the instances confirmed,this model is high in foretelling precision.