Title :
Research of Intrusion Detection Based on Principal Components Analysis
Author_Institution :
Inf. Eng. Inst., Dalian Univ., Dalian, China
Abstract :
The effective way of improving the efficiency of intrusion detection is to reduce the heavy data process workload. In this paper, the dimensionality reduction use of technology in the classic dimensionality reduction algorithm principal component to analysis large-scale data source for reduced-made features of the original data be retained and improved the efficiency of intrusion detection. And use BP neural network training the data after dimensionality reduction, will be effective in normal and abnormal data distinction, and achieved good results.
Keywords :
backpropagation; computer networks; neural nets; principal component analysis; telecommunication security; BP neural network training; PCA; abnormal data distinction; classic dimensionality reduction algorithm; computer network security; data process workload; intrusion detection; large-scale data source; network packet capture feature space dimension; principal component analysis; reduced-made feature; Data engineering; Engines; High-speed networks; Information analysis; Intrusion detection; Nearest neighbor searches; Neural networks; Packaging; Principal component analysis; Space technology; PCA; intrution detection; networksecurity;
Conference_Titel :
Information and Computing Science, 2009. ICIC '09. Second International Conference on
Conference_Location :
Manchester
Print_ISBN :
978-0-7695-3634-7
DOI :
10.1109/ICIC.2009.36