Application of intrusion detection technology in network safety based on machine learning

Abstract

As an important means to ensure network safety, intrusion detection technology can be much more efficient by introducing machine learning. The present paper proposes a machine learning method for intrusion information detection, which can fully exploit the envelope advantages of Elman neural network and the advantages of robust SVM noise data elimination, and can then combine the two to solve the safety risks of intrusion detection of information systems to ensure the safety of information systems. Elman neural network intrusion detection clusters the text of the network packet by clustering algorithm, which largely reduces the defect of missing text information. It also improves the ability to detect abnormal behaviour between network packet sequences. Meanwhile, robust SVM neighbour classification intrusion detection can achieve the feature space weighting of the optimal classification face host system log, eliminate the negative impact of noise data, reduce the false alarm rate of intrusion detection, and improve the detection accuracy. The result shows that when the false alarm rate is 0, the intrusion detection rate based on robust SVM neighbour classification can achieve 87.3%; when the false alarm rate is 2.8%, the detection rate can reach as high as 100%.