Authors:
A.S. Vignesh Raja, K. Daniel Jasper, Rasha Aljaafreh, S.K. Yogeshwarran, Muhammad Saleem

Addresses:
1,2,4Department of Computer Science and Engineering, SRM Institute of Science and Technology, Ramapuram, Chennai, Tamil Nadu, India. 3Department of Computer Science, University of Technology Sydney, Ultimo, Australia. 5Department of Management Science and Engineering, Kunming University of Science and Technology, Kunming, China.   ar6256@srmist.edu.in1, dk9127@srmist.edu.in2, raljaafreh@gmail.com3, sy4289@srmist.edu.in4, m.saleem647@gmail.com5  

Abstract:

This article examines how blockchain-based DApps and federated learning can improve data management, privacy, and collaborative machine learning. In an age of exponential technological innovation, secure and decentralised data management is essential. Blockchain technology offers hope with a decentralised, immutable record that assures transparency, security, and trust without intermediaries. Our research explores DApps' complex architecture, protocols, and cryptographic processes, as well as their potential uses and influence across sectors. We also examine federated learning, a pioneering privacy-preserving machine learning method. Federated learning allows collaborative model training across dispersed devices or servers without data aggregation, protecting data. We evaluate federated learning systems' performance, scalability, and privacy across varied datasets and tasks through rigorous testing and review. The results show that dataset properties should be used to choose model architectures and training configurations and that privacy-preserving strategies can reduce privacy leaks. Federated learning's scalability and resource efficiency could revolutionise distributed collaborative machine learning, according to our findings. This comprehensive examination illuminates the complex relationship between decentralized computing, cryptographic innovation, and blockchain and federated learning's promise to create a more robust, transparent, and decentralised digital economy. 

Keywords: Federated Learning; Decentralized Applications; Convolutional Neural Networks; Blockchain-Based Decentralized Applications; Reshaping Data Management; Blockchain Technology; Machine Learning.

Received on: 15/04/2023, Revised on: 28/07/2023, Accepted on: 11/10/2023, Published on: 22/12/2023

FMDB Transactions on Sustainable Computer Letters, 2023 Vol. 1 No. 4, Pages: 228-240

• Views : 210
• Downloads : 19

Download PDF