Leveraging Transfer Learning to Optimize Edge Computing in Resource-Constrained Settings.

EasyChair Preprint no. 11823

Abstract

Edge computing is a distributed computing paradigm that brings computation and data storage closer to the edge devices, enabling real-time and low-latency processing. Transfer learning, with its ability to leverage pre-trained models, can play a crucial role in enhancing machine learning applications in edge computing environments. This paper explores the challenges and opportunities of applying transfer learning in edge computing scenarios.  We discuss the considerations for model selection, training, and deployment in resource-constrained edge devices. Additionally, we explore techniques for efficient knowledge transfer, model compression, and federated learning to optimize the performance and energy efficiency of edge devices. Our findings demonstrate the potential of transfer learning to enable intelligent applications at the edge with limited computational resources.

Keyphrases: Edge Computing, Federated Learning, Model Compression, model selection, resource-constrained devices, Transfer Learning

@Booklet{EasyChair:11823,
  author = {Asad Ali and Muzamil Abbas},
  title = {Leveraging Transfer Learning to Optimize Edge Computing in Resource-Constrained Settings.},
  howpublished = {EasyChair Preprint no. 11823},

  year = {EasyChair, 2024}}