针对现有BP神经网络算法存在的平坦曲面和局部极小点问题,提出了极小点跃迁遍历BP算法,以取得更好的收敛效果.将改进的BP算法应用于无线局域网入侵检测系统.实验表明,改进的算法提高了入侵检测的准确性和实时性.
In view of the existence of slightly curved surfaces and local minima problems with the application of the BP neural network algorithm,ergodic BP algorithm based on minimal point escape was proposed to achieve better convergence effect.The improved BP algorithm was applied to the wireless network intrusion detection system.Experiments show that,the improved BP algorithm enhances the accuracy and real time of intrusion detection systems.