Quanzhou Xiang

Title: Quanzhou Xiang: Innovator in Machine Learning

Introduction

Quanzhou Xiang is a prominent inventor based in Shanghai, China. He has made significant contributions to the field of machine learning, particularly in developing systems that enhance the robustness of machine learning models against adversarial attacks. His innovative approach addresses critical challenges in data imbalance and sophisticated fraud detection.

Latest Patents

Quanzhou Xiang holds 1 patent for his invention titled "Systems and methods for machine learning using adversarial learning and data transformation." This patent discloses a machine learning hardening system and method designed to protect machine learning models from adversarial attacks. The method involves preprocessing data for logistic regression and augmenting the dataset with synthetic fraudulent transactions generated by a Negative-Calibrated Generative Adversarial Network (NCGAN) to ensure balance. Additionally, adversarial behavior is modeled to simulate fraudster responses to detection models, employing a Stackelberg game framework. A logistic regression model is trained using a sum of top-K loss function to focus on the transactions with the greatest loss, thereby enhancing the detection of fraudulent activity.

Career Highlights

Quanzhou Xiang is currently employed at HSBC Software Development (Guandong) Limited, where he continues to innovate in the field of machine learning. His work is characterized by a strong focus on developing solutions that address real-world challenges in data security and fraud detection.

Collaborations

Quanzhou has collaborated with notable colleagues, including Bing Zhu and Ziyuan Li, who share his commitment to advancing technology in machine learning.

Conclusion

Quanzhou Xiang's contributions to machine learning and his innovative patent demonstrate his dedication to enhancing the security and effectiveness of machine learning models. His work is paving the way for more robust systems in the fight against fraud and data imbalance.