FedHealth 2: Weighted Federated Transfer Learning via Batch Normalization for Personalized Healthcare
It addresses data privacy and domain shift issues for personalized healthcare applications, representing an incremental extension of prior work.
The paper tackles domain shifts and personalization in federated learning for healthcare by proposing FedHealth 2, which uses client similarities and weighted averaging with local batch normalization, resulting in over 10% accuracy improvement in activity recognition and effective personalized models.
The success of machine learning applications often needs a large quantity of data. Recently, federated learning (FL) is attracting increasing attention due to the demand for data privacy and security, especially in the medical field. However, the performance of existing FL approaches often deteriorates when there exist domain shifts among clients, and few previous works focus on personalization in healthcare. In this article, we propose FedHealth 2, an extension of FedHealth \cite{chen2020fedhealth} to tackle domain shifts and get personalized models for local clients. FedHealth 2 obtains the client similarities via a pretrained model, and then it averages all weighted models with preserving local batch normalization. Wearable activity recognition and COVID-19 auxiliary diagnosis experiments have evaluated that FedHealth 2 can achieve better accuracy (10%+ improvement for activity recognition) and personalized healthcare without compromising privacy and security.