摘要
网络安全是当前网络管理领域研究中的重点,针对BP神经网络的阈值和连接权值优化问题,提出一种群智能算法优化BP神经网络参数的方法,并将其应用于网络安全。首先对群智能算法中的生物地理学算法进行改进,加快其收敛速度,然后采用改进生物地理学算法择BP神经网络的阈值和连接权值,最后采用网络入侵数据集对其有效性和优越性进行测试。结果表明,生物地理学算法可以快速找到BP神经网络的最优阈值和连接权值,提高了网络入侵检测的正确率,可以有效的保护网络系统的安全。
Network security is the focus research in the field of network management at present,according to the BP neural network's threshold and connection weights optimization problems,this paper put forward a BP neural network which parameters are optimizing by swarm intelligence algorithm and are applied in network security.Firstly,the biogeography-based optimization algorithm is improved to speed convergence speed,and then biogeography-based optimization algorithm is used to select the threshold and connection weights of BP neural network,finally the network intrusion data set are used to test its validity and superiority.The results show that,the biogeography algorithm can quickly find the optimal he threshold and connection weights of BP neural network,improved network intrusion detection rate,and can protect the safety of network system.
出处
《激光杂志》
北大核心
2015年第4期143-146,共4页
Laser Journal
基金
广东省科技厅研究项目(132300410479)
关键词
网络安全
生物地理优化算法
神经网络
网络入侵
network security
biogeography-based optimization algorithm
neural network
network intrusion
