Kun-yi Chen

Title: Kun-yi Chen: Inventor in Federated Learning

Introduction

Kun-yi Chen is an inventor based in Taichung, Taiwan. His work includes a patented approach to multi-layer federated learning that uses distributed clustering to organize participating nodes.

Latest Patents

Chen is associated with 1 patent: “Computer Implemented Multi-Layer Federated Learning Method Based on Distributed Clustering and Non-Transitory Computer Readable Medium Thereof.” The method measures feature-distribution similarity among nodes with non-independent and identically distributed data, then groups the nodes into clusters. A federated learning algorithm updates the local models within each cluster, and a multi-layer aggregation mechanism combines their contributions until the clustering results meet the required criteria. The described system also uses blockchain-based components, including model aggregation, an API, time-series synchronization, and the InterPlanetary File System (IPFS). The patent states that this architecture can improve learning performance.

Career Highlights

Chen is associated with Asia University. His patent focuses on coordinating distributed data and model updates in federated learning, a method that enables collaborative model training without centralizing participating nodes’ data.

Collaborations

Chen’s listed coworkers include Zon-yin Shae and Jing-pha Tsai.

Conclusion

Kun-yi Chen’s patent presents a clustered, multi-layer approach to federated learning, combining distributed model updates with blockchain-based system components.