Affiliations: School of International Education, Guizhou Normal University, Yunyan District, Guiyang, Guizhou 550001, China | E-mail: gpcguip@yeah.net
Correspondence:
[*]
Corresponding author: School of International Education, Guizhou Normal University, Yunyan District, Guiyang, Guizhou 550001, China. E-mail: gpcguip@yeah.net.
Abstract: The security precaution of enterprise information resource database is a problem that enterprises pay close attention to extensively. In this paper, intrusion detection technology was studied, and a hybrid genetic algorithm which combined genetic algorithm with Back Propagation (BP) neural network was developed. The algorithm was tested using KDD CUP 99 data set. The results showed that the convergence effect of the hybrid genetic algorithm was good, the detection rate of the algorithm for different attacks was higher than 80%, and the accuracy rate was over 90%. The detection rate, false alarm rate, accuracy rate and detection time of the hybrid genetic algorithm were 91.36%, 6.72%, 92.24%, and 0.34 s respectively, suggesting a better detection performance. The hybrid genetic algorithm also had an accuracy rate of 98.42% in the practical application in the information resource database of an enterprise in Guizhou, China. The hybrid genetic algorithm developed in this study has a good performance in intrusion detection and has great values for the security protection of enterprise information resource database.
Keywords: Genetic algorithm, big data, network security, intrusion detection
