Liefeng Bo

Title: Liefeng Bo and Asynchronous Federated Learning

Introduction

Liefeng Bo is an inventor based in Mountain View, California. His patent work includes a system for training federated learning models asynchronously, an approach designed to coordinate model updates across multiple institutional clients.

Latest Patents

Bo is associated with 1 patent: “System and Method for Asynchronous Multi-Aspect Weighted Federated Learning.” The invention describes a master system that coordinates K institutional clients through an aggregation thread and separate client management threads. Each client management thread instructs its client to train a local federated learning model and sends the resulting training event to the aggregation thread. The aggregation thread updates the global model and returns an aggregation event to the relevant client management thread, enabling asynchronous training and aggregation.

Career Highlights

Bo is associated with Beijing Wodong Tianjun Information Technology Co., Ltd. His patent reflects work on federated learning systems that coordinate local model training and global model updates across institutional clients.

Collaborations

Huasong Shan and Tao Zeng are identified as Bo’s coworkers.

Conclusion

Liefeng Bo’s patent focuses on asynchronous coordination in federated learning, linking local client training with updates to a shared global model.