A Hybrid Ensemble Deep Learning Models to Enhance the Cloud Security to Mitigate the DOS Attacks

B. Vinothkumar, M. Dharani, M. Udhayakumar, Sowmiya S., Gowtham Kumar B., Naveen R.

2025

Abstract

Aim: Enhancing cloud security through the development of a hybrid ensemble of deep learning models to efficiently identify and counteract Denial of Service (DoS) assaults is the main goal of this research. Materials and Method: In this research, there are two groups.: Group 1 (LSTM) and Group 2 (CNN) of 26 samples each with a G Power of 80%, a threshold of 0.05, and a 95% confidence interval. Result: The CNN model outperformed the LSTM model in accuracy, 92.56% to 96.74%, while the LSTM model ranged between 85.42% to 91.87%. In addition, CNN had lower false positive rates ranging from 2.87% to 4.14% compared to LSTM, which had 4.32% to 6.89%. CNN also had a better stability, with a standard deviation of 1.6743, whereas LSTM had 2.8567. Conclusion: These results confirm the effectiveness of CNN in DoS detection, consistent with studies on cloud security and AI-based threat detection.

Paper Citation

in Harvard Style

Vinothkumar B., Dharani M., Udhayakumar M., S. S., B. G. and R. N. (2025). A Hybrid Ensemble Deep Learning Models to Enhance the Cloud Security to Mitigate the DOS Attacks. In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25; ISBN 978-989-758-777-1, SciTePress, pages 242-248. DOI: 10.5220/0013911200004919

in Bibtex Style

@conference{icrdicct`2525,
author={B. Vinothkumar and M. Dharani and M. Udhayakumar and Sowmiya S. and Gowtham B. and Naveen R.},
title={A Hybrid Ensemble Deep Learning Models to Enhance the Cloud Security to Mitigate the DOS Attacks},
booktitle={Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25},
year={2025},
pages={242-248},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013911200004919},
isbn={978-989-758-777-1},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25
TI - A Hybrid Ensemble Deep Learning Models to Enhance the Cloud Security to Mitigate the DOS Attacks
SN - 978-989-758-777-1
AU - Vinothkumar B.
AU - Dharani M.
AU - Udhayakumar M.
AU - S. S.
AU - B. G.
AU - R. N.
PY - 2025
SP - 242
EP - 248
DO - 10.5220/0013911200004919
PB - SciTePress