基于支持向量机补偿的灰色模型网络流量预测

摘要

针对网络测量与控制技术中提高流量预测准确性的问题,提出基于支持向量机残差补偿的灰色模型网络流量预测模型.该模型采用灰色模型进行趋势预测,支持向量机进行残差序列预测,实现残差补偿.实验结果表明:该模型具有预测模型样本小,预测精度高等优点,适合于网络流量预测.%In order to improve flow prediction precision in network measurement and control technology, a method of grey model and support vector machines regression was used for traffic flow prediction. The trend was forecasted by grey model. Residual sequence was forecasted by SVM. Predictions were added to the result of grey model of network traffic. This method could modify grey model of network traffic. Experiment results showed the modle was of small forecast model sample and high accuracy, and could effectively predict the network traffic.