In order to improve the security of the network

this paper proposes an improved support vector machine algorithm for network intrusion detection.Firstly

kernel principal component analysis is used to extract data network feature

accelerate the network intrusion detection rate

then the particle swarm optimization algorithm is used to optimize the support vector machine parameters

improve the network detection accuracy rate.The simulation results show that

the proposed method has improved network intrusion detection rate

reduce the failure rate

and speed the network intrusion detection

it is an efficient

real-time network intrusion detection algorithm.