基于改进天牛群算法优化的BP神经网络的入侵检测

针对传统BP神经网络的入侵检测中,BP神经网络模型存在容易陷入局部最优、收敛速度慢、初始值随机性较大等缺点,本文提出改进天牛群算法(Beetle Swarm Optimization,BSO)用于优化BP神经网络的权值与阈值,并采用可变的感知因子及导向性的学习策略,以增强算法跳出局部最优的能力,提升算法全局寻优能力。利用天牛群算法群体智能的特点,提高BP神经网络的收敛速度。并将天牛群优化的BP神经网络模型应用于入侵检测,仿真实验结果表明优化后的BP神经网络模型能够显著提高模型的收敛速率和对入侵数据的检测率,降低误报率。

Intrusion Detection Based on Improved BP Neural Network Based on Improved Beetle Swarm Optimization

In the intrusion detection of traditional BP neural network, BP neural network model is easy to fall into local optimum,slow convergence rate and large initial value randomness.This paper proposes to improve the Beetle Swarm Optimization(BSO) algorithm. It is used to optimize the weights and thresholds of BP neural network,and adopts variable perceptual factors and guiding learning strategies to enhance the ability of the algorithm to jump out of local optimum,improve the global optimization ability of the algorithm,and use the herd algorithm for group intelligence.Features to improve the convergence rate of BP neural networks.The BP neural network model optimized by Tianniu Group is applied to intrusion detection.The simulation results show that the optimized network model can significantly improve the convergence rate of the algorithm model and the detection rate of intrusion data,and reduce the false positive rate.