Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{203725,
        author = {Eknath P. Borde and Pratik V. Kshirsagar and Paras P. Medpalliwar and Saurav N. Shende and Mayur P. Barase and Prof. Anand Donald},
        title = {Cloud Intrusion Detection Using Co-operative Machine Learning Technique},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {1},
        pages = {1113-1119},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=203725},
        abstract = {Traditional Intrusion Detection Systems (IDS) rely on centralized data structures. This creates major risks for user privacy, leads to significant communication bandwidth issues, and poses single-point-of-failure (SPOF) threats. Additionally, real-world network traffic data often suffers from extreme class imbalance. This skews conventional machine learning (ML) models, causing them to overlook rare, stealthy zero-day and infiltration attacks. This paper introduces a decentralized Cloud-Federated Intrusion Detection System that uses cooperative machine learning to train Deep Neural Networks (DNN) locally across distributed edge nodes. Our system operates on a Google Cloud Platform (GCP) coordinator layer and utilizes the Federated Averaging (FedAvg) algorithm to iteratively combine localized model weights, keeping raw network packets out of the cloud. To address severe class imbalance within edge nodes, we incorporate the Synthetic Minority Over-sampling Technique (SMOTE) locally before training. Rigorously tested against the CSE-CIC-IDS2018 benchmark dataset, the proposed architecture achieves a global accuracy of 86.78% and notable improvements in recall. It reaches a 100% detection rate for highly evasive WebAttacks and an 86% detection rate for stealthy Infiltration attacks.},
        keywords = {Intrusion Detection System, Federated Learning, Cooperative Machine Learning, SMOTE, Deep Neural Network, CSE-CIC-IDS2018.},
        month = {June},
        }