Evaluation Mechanism for Decentralized Collaborative Pattern Learning in Heterogeneous Vehicular Networks

Collaborative machine learning, especially Federated Learning (FL), is widely used to build high-quality Machine (ML) models in the Internet of Vehicles (IoV). In this paper, we study performance evaluation problem an inherently heterogeneous IoV, where final across network are not identical and computed on different standards. Previous studies assume that local agents receiving data from same phenomenon, a model fitted them. However, “one fits all” approach leads biased individual agents. We propose general measure agents, common knowledge correlation between explored. Experimental results indicate our scheme efficient these settings.