A recent work I co-authored, entitled “Balancing Federated Learning Trade-Offs for Heterogeneous Environments”, investigated trade-offs associated with Federated Learning (FL) related to system and data/statistical heterogeneity. More specifically, the work aims to acquire initial insights for simple strategies to balance decisions for how local training is done on high-power and low-power edge devices in a FL process.

This work as accepted for publication through the 2023 IEEE International Conference on Pervasive Computing and Communications (PerCom) Work-in-Progress (WiP) session. At the conference’s banquet on Thursday, March 16, 2023, it was awarded the Best Paper Award for the WiP session of PerCom 2023.