US Patent Application 18027371. ELECTRONIC DEVICE AND METHOD FOR FEDERATED LEARNING simplified abstract
Contents
ELECTRONIC DEVICE AND METHOD FOR FEDERATED LEARNING
Organization Name
Inventor(s)
ELECTRONIC DEVICE AND METHOD FOR FEDERATED LEARNING - A simplified explanation of the abstract
This abstract first appeared for US patent application 18027371 titled 'ELECTRONIC DEVICE AND METHOD FOR FEDERATED LEARNING
Simplified Explanation
- The present disclosure describes an electronic device and method for federated learning. - The electronic device is designed to perform federated learning at a central processing apparatus. - The device includes a processing circuitry that is configured to determine a group of distributed nodes for generating a global model parameter. - The group of distributed nodes is selected based on the correlation between their local training data, which must meet a specific correlation requirement. - The device further generates the global model parameter based on the local model parameters of the group of distributed nodes. - The local model parameters are generated by each distributed node based on their respective local training data. - This approach allows for collaborative learning without sharing raw data, as the global model parameter is generated based on the local model parameters of the distributed nodes. - The method enables efficient and privacy-preserving federated learning by leveraging the correlation between local training data.
Original Abstract Submitted
The present disclosure provides an electronic device and a method for federated learning. The electronic device for federated learning at a central processing apparatus comprises a processing circuitry which is configured to: determine a group of distributed nodes for generating a global model parameter among a plurality of distributed nodes, wherein the correlation between local training data of the group of distributed nodes meets a specific correlation requirement; and generate the global model parameter based on local model parameters of the group of distributed nodes, wherein the local model parameters are generated by the group of distributed nodes based on respective local training data thereof.