TY  - JOUR
TI  - An Hybrid Intrusion Detection Approach based on SVM Classification and k-NN
AU  - A. Anbarasa Kumar
AU  -  Kumar Parasuraman
JO  - International Journal of Scientific Research in Computer Science, Engineering and Information Technology
PB  - Technoscience Academy
DA  - 2018/06/30
PY  - 2018
DO  - https://doi.org/10.32628/IJSRCSEIT
UR  - https://ijsrcseit.com/CSEIT1835200
VL  - 3
IS  - 5
SP  - 839	
EP  - 852
AB  - Communication of information between various organizations to maintain a high level security to ensure safe and trusted communication is very important. Nowadays in internet secure data communication is not may be possible and other network also. There is  thread of intrusion and misuses  are occurs in any kinds of networks. We need to detect and recognize these threads and prevent cyber attacks. In this paper IDS (Intrusion Detection System) using a SVM classifier (Support Vector Machine) and   to prevent the network attacks like probe attacks , DoS denial of service, R 2 L remote to user ,U 2 R user to root  attacks  using   SSP (Sniffer and Snooping Process). Intrusion Detection has been an essential countermeasure to secure registering frameworks from noxious attacks. To enhance identification execution and decrease predisposition towards visit attacks, this paper proposes a hybrid strategy in view of SVM classification and k-NN procedure. Trial comes about show that the proposed strategy beats baselines regarding different assessment criteria. Specifically, for U2R and R2L attacks, the F1-scores of the proposed technique are substantially higher than those of baselines. Besides, comparisons with some ongoing hybrid approaches are additionally recorded. The outcomes show that the proposed strategy is focused.