Yan Xu

Title: Innovations of Yan Xu in Transaction Composition Graph Node Embedding

Introduction

Yan Xu is an accomplished inventor based in Richmond Hill, Canada. He has made significant contributions to the field of computer science, particularly in the area of transaction composition graph node embedding. His innovative work has led to the development of a unique patent that enhances data processing and analysis.

Latest Patents

Yan Xu holds a patent for a process titled "Transaction Composition Graph Node Embedding." This computer-implemented process involves traversing a data flow of transactions to convert a full graph into multiple directed acyclic subgraphs or paths in spanning trees. The method takes one-by-one nodes as input to a predetermined neural network, generating a set of one-hot vectors for all nodes. It computes an embedding vector from a corresponding one-hot vector, calculates the probability that an output node is nearby, and embeds the node to a latent feature vector. This innovative approach has the potential to improve the efficiency of transaction analysis.

Career Highlights

Yan Xu is currently employed at International Business Machines Corporation (IBM), where he continues to push the boundaries of technology and innovation. His work at IBM allows him to collaborate with some of the brightest minds in the industry, contributing to groundbreaking advancements in computer science.

Collaborations

Throughout his career, Yan Xu has worked alongside talented colleagues, including Yan Luo and Liujia Shao. These collaborations have fostered an environment of creativity and innovation, leading to the development of cutting-edge technologies.

Conclusion

In summary, Yan Xu's contributions to the field of transaction composition graph node embedding exemplify the spirit of innovation. His patent and work at IBM highlight his commitment to advancing technology and improving data processing methods.