Title :
Study on improvement of recognition ability of intrusion detection system
Author :
Yanwei, Fu ; Yingying, Zhu
Author_Institution :
Network Center, Changzhou Univ., Changzhou, China
Abstract :
This paper integrates PCA method and optimized LMBP neural network into intrusion detection system. The feasibility and efficiency of the optimized algorithm are approved by utilizing the KDDCUP99 dataset, constructing suitable training samples and testing samples and using MATLAB simulation experiment. Experiment shows that the improved algorithm has obvious superiority in training times and accuracy.
Keywords :
backpropagation; neural nets; principal component analysis; security of data; KDDCUP99 dataset; MATLAB simulation; PCA method; intrusion detection system; optimized LMBP neural network; principal component analysis; recognition ability improvement; Computational modeling; Databases; Jacobian matrices; Optimization; Intrusion Detection; Levenberg-Marquardt(LM) Algorithm; Neural Network; Principal Component Analysis;
Conference_Titel :
Communication Technology (ICCT), 2010 12th IEEE International Conference on
Conference_Location :
Nanjing
Print_ISBN :
978-1-4244-6868-3
DOI :
10.1109/ICCT.2010.5689069
