Network intrusion detection by rough set and least squares support vector machine
Duan Xian-hui, Liu Zhiguo, Hua Liu · 2010
The hybrid method of rough set and least squares support vector machine is presented to network intrusion detection in the paper. The 460 experimental data in KDDCUP99 are employed to research the proposed detection model. In the experimental data, 300 is the number of normal data, and the number of four fault types: DoS, R2L, U2R and Probe is 40 respectively. The experimental results show that the detection accuracy of RS-LSSVM is superior to SVM and BPNN.