EUVIMEDEUVIMED entdecken
Crossref

A Hybrid Machine Learning Framework for Real-Time Cyber Threat Prediction and Adaptive Security Decision-Making in IoT-Cloud Environments

Dr. Araddhana Manisha Arvind Deshmukh, Dr Gayatri S Pandi, Dr Namita Kale, Musthafa Kamal Pasha A, A Hency Juliet, Mr. M. Ramesh

Computer and Decision Making An International Journal · 15.11.2026

DOI: 10.1922/comdem.v4i1.19165 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Computer and Decision Making An International Journal, online erschienen am 15.11.2026. Verfasst von Dr. Araddhana Manisha Arvind Deshmukh und 5 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The increasing integration of Internet of Things devices with cloud environments has created growing cybersecurity challenges requiring rapid and reliable threat detection. This study developed a hybrid machine learning framework for cyber threat prediction and adaptive security decision-making in IoT-cloud environments. Network traffic characteristics, system-resource indicators, and behavioural variables were analysed to distinguish normal from malicious activity. Logistic Regression, Decision Tree, and Random Forest models were developed and combined using a majority-voting approach. Model performance was evaluated using accuracy, precision, recall, F1-score, and confusion matrix analysis. The findings showed clear differences between normal and attack traffic, particularly in packet activity, connection frequency, failed login attempts, CPU and memory usage, cloud latency, and anomaly score. Logistic Regression achieved the highest predictive performance, while Random Forest and the hybrid model also demonstrated strong classification capability. The hybrid approach correctly classified most testing observations and supported reliable identification of malicious network behaviour. An adaptive security decision layer further linked threat severity with practical responses such as monitoring, access restriction, blocking, and isolation. Overall, the findings indicate that simple and interpretable machine learning approaches can provide effective support for cyber threat detection and adaptive security management in IoT-cloud environments.

Zitieren

APA

Deshmukh, D. A. M. A., Pandi, D. G. S., Kale, D. N., A, M. K. P., Juliet, A. H., & Ramesh, M. M. (2026). A Hybrid Machine Learning Framework for Real-Time Cyber Threat Prediction and Adaptive Security Decision-Making in IoT-Cloud Environments. Computer and Decision Making An International Journal. https://doi.org/10.1922/comdem.v4i1.19165

Vancouver

Deshmukh DAMA, Pandi DGS, Kale DN, A MKP, Juliet AH, Ramesh MM. A Hybrid Machine Learning Framework for Real-Time Cyber Threat Prediction and Adaptive Security Decision-Making in IoT-Cloud Environments. Computer and Decision Making An International Journal. 2026. doi:10.1922/comdem.v4i1.19165

BibTeX
@article{Deshmukh2026,
  author = {Deshmukh, Dr. Araddhana Manisha Arvind and Pandi, Dr Gayatri S and Kale, Dr Namita and A, Musthafa Kamal Pasha and Juliet, A Hency and Ramesh, Mr. M.},
  title = {A Hybrid Machine Learning Framework for Real-Time Cyber Threat Prediction and Adaptive Security Decision-Making in IoT-Cloud Environments},
  journal = {Computer and Decision Making An International Journal},
  year = {2026},
  doi = {10.1922/comdem.v4i1.19165}
}
Kostenlos starten →