Enhancing Network Security in Smart AI Environments: Innovative Approaches to Detecting Anomalies
Neerudu Anusha, Thukkani Swetha, Shreyas Shrenik Paraj Patil, Priyanka, Shailendra Singh Sikarwar, Dungar Singh · 2024
Network intrusion poses a significant threat to the security of modern network communications, necessitating the development of effective intrusion detection systems (IDS) to safeguard network services and data. Machine learning techniques have become strong tools for identification of suspicious patterns in network traffic. This paper proposes ensemble learning using both SVMs and the Random Forest for network intrusion detection. The performances of the proposed methods are evaluated using the enhanced version of the KDDCUP’99 dataset, known as the NSL-KDD dataset. The main focus of our intrusion detection algorithm is to classify incoming network traffic as normal or malicious based on the given 41 features describing each pattern of the traffic. The accuracy in our approach is far above 95%, with the best classifier being Random Forest, which outperforms SVMs. This is an outstanding achievement in smart environment network security and privacy.