摘要
针对传统的BP神经网络所具有的算法收敛速度慢、容易陷入局部极小等固有缺陷,文章提出了一种优化改进算法的BP神经网络,并利用其构建了分布式入侵检测模型。通过仿真实验证明,该入侵检测模型不仅具有良好的全局搜索能力,而且算法收敛速度较快,具有极佳的入侵检测效果。
Research found that the traditional BP neural networks, the algorithm adopted by the slow convergence speed and easy to fall into local minimum inherent defects. This paper proposes a optimization algorithm for BP neural networks, and create a distributed intrusion detection model. Through the simulation experiments show that this model not only has good global search ability, and the algorithm convergence speed faster, intrusion detection has excellent effect.
出处
《无线互联科技》
2016年第3期11-12,28,共3页
Wireless Internet Technology
关键词
BP神经网络
分布式入侵检测模型
构建
BP neural networks
distributed intrusion detection model
construction