TY  - JOUR
AU  - Dashdondov, Khongorzul
AU  - Chamazkoti, Mahjoobe Nazari
AU  - Abdusalomov, Akmalbek
AU  - Ullah, Habib
AU  - Khan, Muhammad Zubair
AU  - Ali, Bakht Sher
PY  - 2026
DA  - 2026/01/12
TI  - Hybrid XGBoost-CNN Model for Anomaly Detection: A New Approach for IoT Wireless Sensor Networks
JO  - ICCK Transactions on Advanced Computing and Systems
T2  - ICCK Transactions on Advanced Computing and Systems
JF  - ICCK Transactions on Advanced Computing and Systems
VL  - 2
IS  - 1
SP  - 42
EP  - 52
DO  - 10.62762/TACS.2025.354651
UR  - https://www.icck.org/article/abs/TACS.2025.354651
KW  - IoT wireless sensor networks
KW  - anomaly detection
KW  - XGBoost
KW  - convolutional neural networks
KW  - feature selection
KW  - hybrid model
KW  - real-time detection
AB  - The Internet of Things (IoT) continues to expand rapidly, resulting in increasingly heterogeneous and complex wireless sensor networks (WSNs). Traditional anomaly detection approaches cannot cope with dynamic traffic patterns, high data volumes, and strict resource constraints. This study presents a hybrid XGBoost-CNN model that integrates XGBoost-based feature selection with a lightweight Convolutional Neural Network optimized for IoT environments. The proposed model was evaluated using real-world IoT traffic data and benchmarked against XGBoost, KNN, and SVM. Experimental results show that the hybrid approach improves detection accuracy by up to 2.29%, increases throughput by 8-48%, and reduces computational cost by 3%-8% compared with the baseline models. The model also demonstrated higher energy efficiency under varying attack scenarios. These results indicate that combining the feature selection capabilities of XGBoost with the pattern extraction strength of CNN yields a scalable, accurate, and resource-efficient anomaly detection solution suitable for IoT-WSN devices.
SN  - 3068-7969
PB  - Institute of Central Computation and Knowledge
LA  - English
ER  - 

@article{Dashdondov2026Hybrid,
  author = {Khongorzul Dashdondov and Mahjoobe Nazari Chamazkoti and Akmalbek Abdusalomov and Habib Ullah and Muhammad Zubair Khan and Bakht Sher Ali},
  title = {Hybrid XGBoost-CNN Model for Anomaly Detection: A New Approach for IoT Wireless Sensor Networks},
  journal = {ICCK Transactions on Advanced Computing and Systems},
  year = {2026},
  volume = {2},
  number = {1},
  pages = {42-52},
  doi = {10.62762/TACS.2025.354651},
  url = {https://www.icck.org/article/abs/TACS.2025.354651},
  abstract = {The Internet of Things (IoT) continues to expand rapidly, resulting in increasingly heterogeneous and complex wireless sensor networks (WSNs). Traditional anomaly detection approaches cannot cope with dynamic traffic patterns, high data volumes, and strict resource constraints. This study presents a hybrid XGBoost-CNN model that integrates XGBoost-based feature selection with a lightweight Convolutional Neural Network optimized for IoT environments. The proposed model was evaluated using real-world IoT traffic data and benchmarked against XGBoost, KNN, and SVM. Experimental results show that the hybrid approach improves detection accuracy by up to 2.29\%, increases throughput by 8-48\%, and reduces computational cost by 3\%-8\% compared with the baseline models. The model also demonstrated higher energy efficiency under varying attack scenarios. These results indicate that combining the feature selection capabilities of XGBoost with the pattern extraction strength of CNN yields a scalable, accurate, and resource-efficient anomaly detection solution suitable for IoT-WSN devices.},
  keywords = {IoT wireless sensor networks, anomaly detection, XGBoost, convolutional neural networks, feature selection, hybrid model, real-time detection},
  issn = {3068-7969},
  publisher = {Institute of Central Computation and Knowledge}
}