Policy-Based Federated Learning
This work addresses privacy concerns in federated learning for edge computing applications, though it appears incremental as it builds on existing federated learning methods with policy enforcement.
The paper tackles the challenge of training models with sensitive user data on edge devices by introducing PoliFL, a decentralized framework that supports heterogeneous privacy policies, and demonstrates its ability to perform accurate training and inference within reasonable resource and time budgets on mobile phone use cases like predictive text and image classification.
In this paper we present PoliFL, a decentralized, edge-based framework that supports heterogeneous privacy policies for federated learning. We evaluate our system on three use cases that train models with sensitive user data collected by mobile phones - predictive text, image classification, and notification engagement prediction - on a Raspberry Pi edge device. We find that PoliFL is able to perform accurate model training and inference within reasonable resource and time budgets while also enforcing heterogeneous privacy policies.