摘要
对大数据网络恶意入侵数据的恢复,能够有效保证入侵后网络的正常运行。对入侵数据的准确恢复,需要建立网络恶意入侵数据估计模型,对测试样本进行重构获得系数矩阵,完成数据准确恢复。传统方法更新待恢复区域内的数据,对比迭代前后的网络数据残差与阈值,但忽略了获得样本系数矩阵,导致数据恢复精度偏低。提出基于稀疏编码的大数据网络恶意入侵数据恢复方法。采用遗传优化方法对网络恶意入侵数据的参数进行估计,以数据的对数似然函数作为目标函数,建立网络恶意入侵数据估计模型,将数据样本划分为训练样本和测试样本,采用训练样本对测试样本进行重构获得系数矩阵,采用范数正则化因子惩罚目标函数,使得重构的数据系数矩阵稀疏,由此完成大数据网络恶意入侵数据恢复。实验结果表明,所提方法能够有效地利用相关信息进行数据填补,从而提高破损数据恢复准确度。
This paper proposes a restoration method for malicious intrusion data in big data network based on sparse coding. Firstly, the genetic optimization method is used to estimate parameters of network malicious intrusion data and log - likelihood function of data is taken as objective function to build estimation model of network malicious intrusion data. Then, data samples are divided into training samples and test samples, and the training samples are used to reconstruct test samples, moreover, coefficient matrix is obtained. Finally, the penalty objective function of norm regularization factor is used to make reconstructed data coefficient matrix sparse. Thus recovery of malicious intrusion data in big data network is completed. Simulation results show that the proposed method can effectively use the relevant information to fill data, so as to improve the accuracy of data recovery.
出处
《计算机仿真》
北大核心
2017年第12期279-282,共4页
Computer Simulation
关键词
大数据网络
恶意入侵
数据恢复
Big data network
Malicious intrusion
Data recovery