•  
  •  
 

Abstract

In an increasingly interconnected world, cybersecurity threats have become more sophisticated, necessitating advanced, scalable, and privacy-preserving solutions. MetaGuard emerges as a novel framework that integrates federated learning with hybrid machine learning models, specifically XGBoost and meta-learning, to enhance proactive cyber threat detection. This framework offers a robust, distributed approach to cybersecurity, ensuring high detection accuracy while preserving user privacy through the implementation of differential privacy and homomorphic encryption. MetaGuard leverages distributed nodes to collaboratively train a global model, enabling rapid adaptation to new threats without the need for centralized data aggregation. Experimental evaluations using the CYBER-2024 dataset demonstrate that MetaGuard significantly outperforms traditional centralized models and contemporary federated learning frameworks, with notable improvements in accuracy (by 5%), precision (by 4.7%), recall (by 4.5%), and F1-score (by 4.9%). Despite challenges such as potential computational overhead and communication latency, MetaGuard's scalability and privacy-preserving features establish it as a highly effective solution for addressing modern cybersecurity challenges. Future research will focus on optimizing the federated learning process, integrating additional machine learning techniques, and expanding the framework's applications across various industries.

Reason for Retraction

This article has been retracted at the request of the Editorial Office, following an internal investigation conducted in accordance with the Committee on Publication Ethics (COPE) Retraction Guidelines.

The investigation identified serious concerns affecting the integrity and reliability of the published work. Specifically, one or more of the following issues were confirmed:

  1. Undisclosed use of computer-generated text and/or data, in which substantial portions of the content were produced using algorithmic or artificial intelligence–based tools without transparent disclosure, contrary to the journal's authorship and transparency policies.

  2. Compromised peer-review process, indicating irregularities that undermine the validity, independence or authenticity of the review procedure.

  3. Inappropriate or misleading citations, including references that are irrelevant, improperly used, or appear to artificially inflate citation metrics, thereby distorting the scholarly record.

  4. Authorship-related concerns, including the addition of new author(s) at a later stage of the publication process without adequate justification, documentation, or transparent disclosure, raising unresolved questions regarding author contributions, responsibility, and compliance with the journal's authorship criteria.

The Editorial Office determined that these issues significantly compromise the scientific integrity of the article, and that correction alone would be insufficient to address the concerns. Retraction was therefore deemed necessary to maintain the accuracy and trustworthiness of the scholarly record.

The authors were informed of the findings and the retraction decision. While the authors do not respond to this retraction, the journal has proceeded with the retraction in line with COPE guidance, which permits retraction without author consent when editorial integrity is at risk.

This retraction is issued to alert readers that the findings and conclusions of the article should not be relied upon. The original article will remain accessible for the sake of the scholarly record, but it will be clearly marked as retracted.

Apologies are offered to readers of the journal that this was not detected during the submission process.

Please see the Retraction Notice available at: https://ijcsm.researchcommons.org/ijcsm/vol6/iss3/27

DOI: https://doi.org/10.52866/2788-7421.1372

Download

Share

COinS