TY  - JOUR
T1  - An Evolutionary Support Vector Machine for Intrusion Detection
AU - , Sung-Hae Jun AU - , Kyung-Whan Oh 
JO  - Asian Journal of Information Technology
VL  - 5
IS  - 7
SP  - 778
EP  - 783
PY  - 2006
DA  - 2001/08/19
SN  - 1682-3915
DO  - ajit.2006.778.783
UR  - https://makhillpublications.co/view-article.php?doi=ajit.2006.778.783
KW  - Evolutionary programming
KW  -support vector machine
KW  -intrusion detection
AB  - Today most information interchanges are performed in the internet. In the environment, internet attacks continue to increase. So, the methods used as intrusion detective tools for protecting network systems against diverse attacks are very important. The skills of intrusion are getting more powerful continuously. However, the detection techniques have been hard to catch up with these attacks. Therefore, we need good tools for intrusion detection. Many researches for intrusion detection have been studied. Most of them had a difficulty in classifying intrusions from networks accesses in the case of new patterns of intrusions which were not experienced by predictive models. In this study, we propose an efficient method to settle the problem. Our model is constructed by combining evolutionary programming into support vector machine. This model is able to detect new attacks as well as experienced attacks. We verify an improved performance of our model using KDD Cup-99 task data designed by DARPA.
ER  -