Publication Type : Conference Paper
Publisher : IEEE
Source : 2025 IEEE Silchar Subsection Conference (SILCON)
Url : https://doi.org/10.1109/silcon67893.2025.11327132
Keywords : Accuracy , Machine learning algorithms , Supervised learning , Stacking , Intrusion detection , Artificial neural networks , Network security , Benchmark testing , Reliability engineering , NSL-KDD
Campus : Amaravati
School : School of Engineering
Year : 2025
Abstract : In today’s digital era, due to the fact that persistent cyber threats exist, it is increasingly challenging to protect online activities. Intrusion Detection Systems (IDS) play a vital role within this defense. This paper suggests Artificial Neural Networks (ANN) function as a meta-classifier to increase IDS performance while reducing down false alarm rate. For strengthening the ensemble, the method systematically evaluates the accuracy of multiple base learners, each using different machine learning algorithms. Detection accuracy is improved as false positives get reduced, compared to classifiers of the state-of-the-art, as shown within experiments on benchmark datasets (NSL-KDD, UNSW-NB15, CICIDS2017, and CIC-DDoS2019) by the model proposed.
Cite this Research Publication : Bidyapati Thiyam, Shouvik Dey, Stacked Artificial Neural network (SANN) for Intrusion Detection System, 2025 IEEE Silchar Subsection Conference (SILCON), IEEE, 2025, https://doi.org/10.1109/silcon67893.2025.11327132