Knowledge-Injected Federated Learning

TL;DR

This paper introduces a federated learning framework that incorporates participants' domain knowledge to enhance model training, especially in industry applications, by refining the global model with local knowledge.

Contribution

It proposes a novel federated learning method that injects local domain knowledge into the global model, improving performance in real-world industry scenarios.

Findings

Effective knowledge injection improves model accuracy.

Demonstrated success in industry-level application.

Enhances federated learning with domain expertise.

Abstract

Federated learning is an emerging technique for training models from decentralized data sets. In many applications, data owners participating in the federated learning system hold not only the data but also a set of domain knowledge. Such knowledge includes human know-how and craftsmanship that can be extremely helpful to the federated learning task. In this work, we propose a federated learning framework that allows the injection of participants' domain knowledge, where the key idea is to refine the global model with knowledge locally. The scenario we consider is motivated by a real industry-level application, and we demonstrate the effectiveness of our approach to this application.