摘要
为解决传统网络数据流调度算法的均衡性、冗余和噪音数据造成的存储消耗和学习算法运行效率低下的问题,提出了基于机器学习的复杂网络数据均衡调度算法研究。基于机器学习的数据流特征选取,确定PSH数量标志位,进行数据包大小变换,实现复杂网络的数据流均衡算法。实验数据表明,与传统数据流调度相比,基于机器学习的复杂网络数据流调度均衡性提高,学习算法运行效率提升。
In order to solve the problems of equilibrium,storage consumption caused by redundant and noisy data and inefficient operation of learning algorithms in traditional network data flow scheduling algorithms,a machine learning based data balance scheduling algorithm for complex networks is proposed.Based on the feature selection of data stream in machine learning,the number of PSH flags is determined,and the size of data packets is transformed to realize the data flow equalization algorithm in complex networks.The experimental data show that,compared with traditional data flow scheduling,machine learning-based data flow scheduling in complex networks is more balanced and the learning algorithm is more efficient.
作者
刘鑫
曾铭钰
段幼春
樊贵军
Liu Xin;Zeng Mingyu;Duan Youchun;Fan Guijun(School of Railway Operation and Management,Hunan Railway Professional Technology College,Zhuzhou Hunan 412001,China)
出处
《信息与电脑》
2019年第16期53-54,共2页
Information & Computer
关键词
机器学习
复杂网络
数据流
均衡调度算法
machine learning
complex network
data flow
equilibrium scheduling algorithm
