Personalized Federated Learning with Clustering: Non-IID Heart Rate Variability Data Application
Authors:
Joo Hun Yoo,
Ha Min Son,
Hyejun Jeong,
Eun-Hye Jang,
Ah Young Kim,
Han Young Yu,
Hong Jin Jeon,
Tai-Myoung Chung
Abstract:
While machine learning techniques are being applied to various fields for their exceptional ability to find complex relations in large datasets, the strengthening of regulations on data ownership and privacy is causing increasing difficulty in its application to medical data. In light of this, Federated Learning has recently been proposed as a solution to train on private data without breach of co…
▽ More
While machine learning techniques are being applied to various fields for their exceptional ability to find complex relations in large datasets, the strengthening of regulations on data ownership and privacy is causing increasing difficulty in its application to medical data. In light of this, Federated Learning has recently been proposed as a solution to train on private data without breach of confidentiality. This conservation of privacy is particularly appealing in the field of healthcare, where patient data is highly confidential. However, many studies have shown that its assumption of Independent and Identically Distributed data is unrealistic for medical data. In this paper, we propose Personalized Federated Cluster Models, a hierarchical clustering-based FL process, to predict Major Depressive Disorder severity from Heart Rate Variability. By allowing clients to receive more personalized model, we address problems caused by non-IID data, showing an accuracy increase in severity prediction. This increase in performance may be sufficient to use Personalized Federated Cluster Models in many existing Federated Learning scenarios.
△ Less
Submitted 10 August, 2021; v1 submitted 4 August, 2021;
originally announced August 2021.
