EXPLAINABLE AI IN INTRUSION DETECTION SYSTEMS
1Nikita Nerkar, Atharva Nile, Amit Kulkarni, Onkar Kadlag, Prashant Gadakh
As the use of internet in increasing day by day and the chances of system get compromised due to various types of attacks has increased. Intruders are finding new techniques to compromise the system. The concern about the cyber security is growing and for the user most of the model is perceived as a black box. There is need of finding the attack correctly and then proper reports should be generated to show how the system got compromised. So we are proposing a system where Intrusion Detection System (IDS) can detect the attack and Explainable artificial intelligence tell us about what type of attack is being performed on the system. Intrusion Detection System keeps track of the malicious packets entering in the system. Explainable Artificial Intelligence will show the report on which type of attack took place. In the proposed system we have use the NSL-KDD dataset for classification of attack detected by our proposed Intrusion Detection System.
Intrusion detection System, Explainable artificial intelligence, NSL-KDD, classification