Statistical Feature Fusion Driven Enhanced Network Intrusion Detection

DOI

10.2991/978-94-6239-678-4_22How to use a DOI?

Keywords

Intrusion Detection; Statistical Fusion; Feature Selection; Deep Learning

Abstract

This study presents a feature selection methodology designed to improve the efficacy of Deep Neural Network (DNN)-based intrusion detection systems (IDS). The suggested method uses a statistical fusion strategy that combines variance thresholding and pairwise correlation analysis to find a small but useful set of network traffic features. The method improves model interpretability and performance by concentrating on features that demonstrate significant variability and minimal redundancy. The proposed approach is examined using three established datasets: NSL-KDD, UNSW-NB15, and CIC-IDS2017. Before training the model, one-hot encoding is used to encode categorical features, and standard normalisation is used to make sure that all the features are scaled the same way. The statistical methods create smaller groups of features, which are then used as inputs for a multilayer DNN classifier. Experimental results show that accuracy, precision, recall, F1-score, and false positive rate (FPR) have all improved a lot compared to traditional feature selection methods like Recursive Feature Elimination, Chi-Square, and Random Forest. The suggested method also cuts down on execution time, even when it uses more features in some cases showing that it is more computationally efficient. The results show that the suggested feature selection strategy could greatly improve the performance and reliability of IDS in different network environments.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Ayush Verma
AU  - Manju Khari
PY  - 2026
DA  - 2026/05/28
TI  - Statistical Feature Fusion Driven Enhanced Network Intrusion Detection
BT  - Proceedings of the 2nd International Conference on Recent Advancement and Modernization in Sustainable Intelligent Technologies & Applications (RAMSITA-2026)
PB  - Atlantis Press
SP  - 280
EP  - 291
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6239-678-4_22
DO  - 10.2991/978-94-6239-678-4_22
ID  - Verma2026
ER  -