基于混合PSO小波网络的网络异常检测

提出一种用新型的进化学习算法训练的小波神经网络(WNN).这种新型的进化学习算法是基于粒子群算法(PSO)和共轭下降法(CG)提出的.以往,将粒子群算法用于神经网络的训练一般是可行的.因为粒子群算法相比于其他的优化算法,具有相对简单的结构和快速的收敛速度,然而,由于粒子的搜索坍塌速度过快而导致粒子停滞这种潜在的危险.粒子的持续停滞使搜索结果很难达到全局最优,甚至会陷入局部最优.为了克服粒子群算法缺点提出了改进的混合算法.通过对KDD 99数据集的实验表明,利用新型混合算法训练的小波神经网络对于异常检测具有很高的异常检测率并且又较低的误判率.可见,该方法对于网络异常检测是有效的.

Network anomaly detection based on hybrid PSO warelet networks

A model of wavelet neural network(WNN) using a new evolutionary learning algorithm is proposed in this paper.This new evolutionary learning algorithm is based on a hybrid of Particle Swarm Optimization(PSO) and Conjugate Gradient algorithm(CG).The particle swarm optimizer has previously been used to train neural networks and generally met with success.The advantage of the PSO over many of the other optimization algorithms exists in its relative simplicity and quick convergence.But those particles collapse s...