Distributed training method between terminal and edge cloud server
AssigneeElectronics and Telecommunications Research Institute
InventorsMyung Ki SHIN, Soohwan LEE
Life of this patent
Estimating the term…
Abstract
A distributed training method performed between a terminal and an edge cloud server is disclosed. The distributed training method may include identifying a plurality of user terminals located in a local network coverage of the edge cloud server, providing machine learning models to the plurality of user terminals for training and aggregating results of the training of the machine learning models from the plurality of user terminals when the training of the machine learning models is completed.
Claims18
A distributed training method performed by an edge cloud server, the distributed training method comprising:
- identifying a plurality of user terminals located in a local network coverage of the edge cloud server;
- providing machine learning models to the plurality of user terminals for training;
- and
- aggregating results of the training of the machine learning models from the plurality of user terminals when the training of the machine learning models is completed.
A distributed training method performed by user terminals, the distributed training method comprising:
- receiving machine learning models to be trained from an edge cloud server when the user terminals are located in a local network coverage of the edge cloud server;
- training the machine learning models using a training data set;
- and
- transmitting results of the training of the machine learning models to the edge cloud server when the training of the machine learning models is completed.
An edge cloud server comprising:
- a processor,
- wherein the processor is configured to perform:
- identifying a plurality of user terminals located in a local network coverage of the edge cloud server;
- providing machine learning models to the plurality of user terminals for training;
- and
- aggregating results of the training of the machine learning models from the plurality of user terminals when the training of the machine learning models is completed.
Dependencies are inferred from the claim text.