Improved genetic algorithm to optimize BP neural network for intrusion detection

Intrusion detection is a proactive security technology to defend the attack from internal and external of the network.Intrusion detection based on neural network is commonly used intelligence test method,in which BP neural network is commonly used neural network model.In this paper,BP neural network algorithm easily trapped into local minima and slow convergence problems,the neural network and genetic algorithm combining genetic algorithm with the improved BP neural network weights.Actual results show that the method can effectively improve the training accuracy and the detection rate,shorten the training time.