Distributed training of machine learning models for personalization

Abstract

A user equipment (UE) may include a communication circuit; and at least one processor configured to: obtain user generated data based on user input of a user of the UE; receive, via the communication circuit, training data from a server connected to the UE, wherein the training data includes publicly available data train a machine learning (ML) model based on the user generated data and the training data until a training stop criterion is met, wherein the training stop criterion includes at least one of an achieved convergence of ML models among one or more UEs including the UE, a predetermined ML model quality characteristic value being achieved by the ML model, or an achieved predetermined number of training periods; and transmit, via the communication circuit, the ML model to the server.