Research Article
A review of Federated Learning
@INPROCEEDINGS{10.4108/eai.24-3-2022.2318998, author={Zargar Danish and Ihtiram Raza Khan}, title={A review of Federated Learning}, proceedings={Proceedings of the 3rd International Conference on ICT for Digital, Smart, and Sustainable Development, ICIDSSD 2022, 24-25 March 2022, New Delhi, India}, publisher={EAI}, proceedings_a={ICIDSSD}, year={2023}, month={5}, keywords={machine learning (ml) federated learning (fl) artificial intelligence (ai)}, doi={10.4108/eai.24-3-2022.2318998} }- Zargar Danish
Ihtiram Raza Khan
Year: 2023
A review of Federated Learning
ICIDSSD
EAI
DOI: 10.4108/eai.24-3-2022.2318998
Abstract
With the fast development of Artificial Intelligence (AI) and Machine Learning (ML), data privacy & security is becoming a serious problem. Since the majority of the Machine Learning Models are centralized in nature which requires sharing the private data to train the centralized ML Model i.e., a model which is stored on a central shared server. In this context, a new approach that is decentralized in nature comes for our rescue, this approach is called Federated Learning (FL). [1] [2] Google is the first company to introduce this concept in 2016. [1] Federated Learning is a decentralized ML technique where different devices or clients in a Federated Network train an ML Model at a central shared Server without sharing their private data. Federated Learning preserves data privacy and gives more control to clients over their data. In this paper, I will give a brief introduction to Federated Learning: its classification, applications, Challenges, Security issues & Future Directions.
