An Optimized Ensemble-Based Machine Learning Model for an Intrusion Detection System to Secure IoT Devices

DOI

10.2991/978-94-6239-713-2_17How to use a DOI?

Keywords

IDS; Ensemble Model; Classifiers; DDoS attack; IoT

Abstract

The rapid growth of IoT has significantly increased network traffic, making modern systems more vulnerable to DDoS attacks. Traditional security mechanisms struggle to detect such attacks effectively due to their dynamic and large-scale nature. To address this challenge, this research shows an exhaustive evaluation of ML and DL models for accurate and reliable DDoS attack detection. Five classifiers—Random Forest (RF), XGBoost(XGB), LightGBM (LGBM), Logistic Regression, and NN were implemented and evaluated using a benchmark intrusion detection dataset. The models are evaluated using various matrices such as accuracy, balanced accuracy, precision-score, recall-score, F1-score, ROC–AUC, and training time. Experimental results demonstrate that all models achieve high detection performance, with accuracy exceeding 96%. Among them, the Neural Network model delivers the best overall performance, achieving an accuracy of 99.74%, balanced accuracy of 99.75%, and an F1-score 99.74%, indicating its superior ability to learn complex and non-linear traffic patterns. Gradient boosting models, LightGBM and XGBoost, also exhibit near-perfect detection capability with ROC–AUC values of 1.000 while preserving efficiency with low computational overhead, making them suitable for real-time deployment. In contrast, Logistic Regression and Random Forest show comparatively lower performance due to higher false positive rates and limited representation capacity. The findings confirm that advanced ensemble approaches significantly enhance DDoS detection effectiveness compared to traditional classifiers. This research provides valuable information for selecting appropriate models for intrusion detection systems, particularly in high- speed and IoT-based network-based surroundings, balancing detection accuracy and computational efficiency.

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  - Tejshri N. Shevate
AU  - Sunita Kushwaha
AU  - Balendra Kumar Garg
AU  - R. D. Kumbhar
PY  - 2026
DA  - 2026/06/25
TI  - An Optimized Ensemble-Based Machine Learning Model for an Intrusion Detection System to Secure IoT Devices
BT  - Proceedings of the International Conference on Advances in Computing Technology and Artificial Intelligence (COMPUTATIA 2026)
PB  - Atlantis Press
SP  - 234
EP  - 244
SN  - 2589-4919
UR  - https://doi.org/10.2991/978-94-6239-713-2_17
DO  - 10.2991/978-94-6239-713-2_17
ID  - Shevate2026
ER  -