Shengming Zhang

Title: Innovations of Shengming Zhang in Anomaly Detection

Introduction

Shengming Zhang is an accomplished inventor based in Kearny, NJ (US). He has made significant contributions to the field of anomaly detection through his innovative patent. His work focuses on enhancing the efficiency of transformers for content-aware anomaly detection in event sequences.

Latest Patents

Shengming Zhang holds a patent titled "Efficient transformer for content-aware anomaly detection in event sequences." This patent presents a method for implementing a self-attentive encoder-decoder transformer framework specifically designed for anomaly detection in event sequences. The method involves feeding event content information into a content-awareness layer to generate event representations. It also includes inputting event sequences of two hierarchies into an encoder to capture both long-term and short-term patterns, generating feature maps in the process. During the training stage, a one-class objective is applied to bind the decoded special sequence token with a reconstruction loss for sequence forecasting. In the testing stage, any event representation whose decoded special sequence token lies outside a hypersphere is labeled as an anomaly.

Career Highlights

Shengming Zhang is currently employed at NEC Corporation, where he continues to develop innovative solutions in the field of technology. His expertise in anomaly detection has positioned him as a valuable asset to his organization.

Collaborations

Shengming Zhang collaborates with talented individuals such as Yanchi Liu and Xuchao Zhang, contributing to advancements in their respective fields.

Conclusion

Shengming Zhang's innovative work in anomaly detection showcases his dedication to advancing technology. His patent reflects a significant step forward in the efficient detection of anomalies in event sequences.