Abstract :

Distributed Machine Learning (DML) Enables Efficient Training Over Large-scale Datasets But Also Introduces Significant Security Challenges. Due To Its Decentralized Nature, DML Systems Are More Vulnerable To Malicious Attacks, Particularly Data Poisoning. This Work Addresses These Challenges By Proposing Effective Detection Mechanisms Tailored For Distributed Environments. We Categorize DML Into Two Types: Basic Distributed Machine Learning (Basic-DML) And Semi Distributed Machine Learning (Semi-DML). In Basic-DML, Worker Nodes Independently Process Assigned Data, Whereas In Semi-DML, The Central Server Also Participates In The Training Process Using Additional Computational Resources. To Mitigate Data Poisoning In Basic-DML, A Crosslearning Based Detection Mechanism Is Introduced, Which Enhances The Identification Of Malicious Data. Additionally, A Mathematical Framework Is Developed To Determine The Optimal Number Of Training Iterations For Improved Detection Accuracy. For Semi-DML, An Enhanced Detection Approach Is Proposed That Focuses On Efficient Resource Allocation And Improved Learning Performance. Experimental Observations Demonstrate That The Proposed Techniques Significantly Improve Model Accuracy While Reducing The Impact Of Poisoned Data. Keywords— Distributed Machine Learning, Data Poisoning, Security, Resource Optimization

Published:

07-8-2026

Issue:

Vol. 26 No. 8 (2026)

Page Nos:

796 - 803

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite