基于增量优化极端学习机的网络流量预测模型  

Network traffic forecasting model based on incremental optimization learning machine

作　　者：闫兵[1] 马琰[2] 

机构地区：[1]河南工业职业技术学院图书馆,河南南阳473000 [2]河南工业职业技术学院实验设备处,河南南阳473000

出　　处：《信息技术》2015年第12期190-193,197,共5页Information Technology

摘　　要：为了获得更加理想的网络流量预测结果,针对极端学习机人工设置隐层节点数目的不足,提出一种增量优化极端学习机的网络流量预测模型。首先对极端学习机工作原理和不足进行分析,然后采用增量优化方式提高极端学习机的性能,最后采用具体网络流量时间序列对增量优化极端学习机的性能进行仿真试验。结果表明,相对于其它网络流量预测模型,增量优化极端学习机不仅加快了网络流量建模速度,可以适合于网络流量的长期和在线预测,而且提高了网络流量的预测精度。In order to get more ideal forecasting results of network traffic, in view of manually setting the hidden layers nodes number of extreme learning machine, this paper puts forward a forecasting model for network traffic based on incremental optimization extreme learning machine network traffic. Firstly, the extreme learning machine working principle and shortcomings are analyzed, and then the optimization of incremental mode is adopted to improve the performance of extreme learning machine, finally, the performance of incremental optimization extreme learning machine is test by using specific network traffic time series. The results show that the proposed model not only speeds up the network traffic modeling speed, can be applied to long-term and online prediction for network traffic, but also improves the prediction accuracy of network traffic compared with other network traffic prediction models.

关 键 词：网络流量 极端学习机 增量优化 预测模型 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]