Computational Models for Intelligence and Automation
A Comprehensive Study of Federated Learning Tools and Frameworks for Data-Sensitive Applications
Authors: Manu Narula, Jasraj Meena and Dinesh Kumar Vishwakarma
Publishing Date: 28-08-2025
ISBN: 978-81-975670-1-8
Abstract
Artificial intelligence depends upon various learning techniques to analyze and process data. Most of these techniques store data centrally or clone it to all training devices. While it presents no issue in standard applications, it quickly becomes a potential security risk when sensitive information gets involved in training, including a range of healthcare and finance utilities. Additionally, the data available for training of a Data Sensitive Application is often scattered across isolated data islands, either due to collaborative or competitive factors. Federated Learning (FL) provides a secure and efficient way to analyze such data. Due to the infancy of the technology, many tools and frameworks of FL are not well known and hence may not be used where their potential utility is maximum. This paper analyzes the various implementations and tools of FL and highlights their suitability for different scenarios with the DSAs.
Keywords
Federated Learning, Tools and Frameworks, Internet of Things, Distributed Learning
Cite as
Manu Narula, Jasraj Meena and Dinesh Kumar Vishwakarma, "A Comprehensive Study of Federated Learning Tools and Frameworks for Data-Sensitive Applications", In: Puneet Kumar Gupta (eds), Computational Models for Intelligence and Automation, SCRS, India, 2025, pp. 127-135. https://doi.org/10.56155/978-81-975670-1-8-12