발간년도 : [2023]
| 논문정보 |
|
| 논문명(한글) |
[Vol.18, No.1] A Study of Detection Data Set Using Various Detection Algorithm |
|
| 논문투고자 |
Se Yul Lee |
|
| 논문내용 |
It is virtually impossible to live without digital devices for us today. The importance of information security in the digital environment is becoming an hot-issue every year, like hacking accidents. To prepare for such intrusion, a next-generation firewall (NGFW) including intrusion detection and a next-generation intrusion prevention system to detect security threats of high-speed networks are being released, and systems that can be applied in the cloud environment are also being researched. We are still continuing to research various algorithms to improve the performance of the detection algorithm. Among detection research, various algorithms to evaluate objective detection performance and data sets for learning algorithms are recognized as very important indicators for detection performance. A representative data set is KDD 99 and NSL-KDD of slightly modified. In this paper, three detection algorithms (SVM: Support Vector Machine, FCM: Fuzzy Cognitive Maps, FCM with oriented weight) are applied and tested to compared and analyze the KDD 99 and NSL-KDD data sets. As a result of analysis, through the results of the KDD and NSL-KDD data sets, which greatly affect the three important issues. It is possible to confirm the detects of the KDD data set and to confirm that the detection algorithm is affected. The NSL-KDD evaluation result show that in order to improve the detection rate in dangerous attacks deleting duplicate record or using or selected records has the advantages of improving the attack detection rate. |
|
| 첨부논문 |
|
|
|
|
|
자료실 > 논문지 
