改进极限学习机的网络流量混沌预测
发布时间：2018-12-07 12:42
【摘要】：为了获得更加精确的网络流量预测,降低网络拥塞的频率,提出了改进极限学习机的网络流量预测模型。针对网络流量混沌性分别确定原始网络流量的延迟时间和嵌入维数,采用极限学习机对网络流量的变化特点进行拟合,改进标准学习机,改善学习速度和预测性能,最后通过网络流量数据的预测实验验证其可行性。验证结果表明:与其它网络流量预测模型相比,改进极限学习的网络流量预测结果更加可靠,对网络流量将来变化趋势可以更加准确描述,提高了网络流量预测精度。
[Abstract]:In order to obtain more accurate network traffic prediction and reduce the frequency of network congestion, an improved extreme learning machine network traffic prediction model is proposed. According to the chaos of network traffic, the delay time and embedding dimension of original network traffic are determined, and the characteristics of network traffic change are fitted by extreme learning machine, and the standard learning machine is improved, and the learning speed and prediction performance are improved. Finally, the feasibility of network traffic data prediction is verified by experiments. The results show that compared with other network traffic prediction models, the improved limit learning network traffic prediction results are more reliable, can more accurately describe the future trend of network traffic changes, and improve the accuracy of network traffic prediction.
【作者单位】： 周口职业技术学院信息工程学院;河南应用技术职业学院信息工程学院;周口师范学院计算机科学与技术学院;
【基金】：国家自然科学基金(U1504613) 河南省高校科技创新团队计划(17IRTSTHN009)
【分类号】：TP181;TP393.06
[Abstract]:In order to obtain more accurate network traffic prediction and reduce the frequency of network congestion, an improved extreme learning machine network traffic prediction model is proposed. According to the chaos of network traffic, the delay time and embedding dimension of original network traffic are determined, and the characteristics of network traffic change are fitted by extreme learning machine, and the standard learning machine is improved, and the learning speed and prediction performance are improved. Finally, the feasibility of network traffic data prediction is verified by experiments. The results show that compared with other network traffic prediction models, the improved limit learning network traffic prediction results are more reliable, can more accurately describe the future trend of network traffic changes, and improve the accuracy of network traffic prediction.
【作者单位】： 周口职业技术学院信息工程学院;河南应用技术职业学院信息工程学院;周口师范学院计算机科学与技术学院;
【基金】：国家自然科学基金(U1504613) 河南省高校科技创新团队计划(17IRTSTHN009)
【分类号】：TP181;TP393.06
【相似文献】
相关期刊论文 前10条
1 陈振伟;王茜;黄继红;;基于多神经网络的网络流量预测模型[J];计算机安全;2010年12期
2 于静;王辉;;基于组合模型的网络流量预测[J];计算机工程与应用;2013年08期
3 叶苗;王勇;;基于支持向量回归学习机的网络流量预测[J];桂林工学院学报;2007年02期
4 张正本;翟海庆;;基于组合模型的网络流量预测[J];河南机电高等专科学校学报;2008年06期
5 周丹琪;乔国平;张泉方;;基于K-Factor GARMA模型的网络流量预测[J];江南大学学报(自然科学版);2008年02期
6 姜明;吴春明;张e，
本文编号：2367225
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2367225.html
