Direct Edge-to-Edge Local-Learning-Assisted Model-Based Transfer Learning

With the gradual advancement of technology in the production of consumer goods, the Internet of Things (IoT) systems have experienced rapid development, resulting in a massive amount of data that can be processed using deep neural networks. However, annotating the data and training the models require significant manpower, time, and computational resources. Transfer learning can address this problem. Traditional systems rely on centralized servers for transfer learning. Although several studies have proposed distributed systems for direct edge-to-edge (e2e) instance-based and feature-based transfer learning, they neglect model-based transfer learning (MTL) within the same domain. This leads to lower learning efficiency, lower privacy protection, and higher transmission costs. Therefore, our study proposes direct e2e local-learning-assisted MTL for a direct e2e many-to-many MTL scenario. The method can transfer model structures and weights between distributed devices without relying on powerful centralized servers. The effectiveness of the proposed approach is demonstrated by applying it to various scenarios.