Jiahui Jeff Wang

Title: Jiahui Jeff Wang and Simulation-Augmented Reinforcement Learning

Introduction

Jiahui Jeff Wang is an inventor associated with Dublin, California, and Amazon Technologies, Inc. Her patent work describes machine-learning methods for selecting digital content in real time.

Latest Patents

Wang is credited with 1 patent: “Simulation Augmented Reinforcement Learning for Real-Time Content Selection.” The invention addresses inventory management, where “inventory” means digital space on a provider’s webpage available for delivering content. It uses offline training and simulated runtime conditions to train models that classify content requests based on historical data and determine whether direct or indirect selection is appropriate. For indirect selection, additional reinforcement-learning models can optimize the request. These techniques help providers choose content according to a preferred objective.

Career Highlights

Wang’s listed patent connects her work at Amazon Technologies, Inc. with the application of reinforcement learning to digital content selection. Its offline training approach is designed to prepare models for decisions made in a real-time environment.

Collaborations

Kirtan Modi and Qie Hu Huang are identified as Wang’s coworkers. The patent information provided does not specify their individual roles in the invention.

Conclusion

Jiahui Jeff Wang’s patent presents a simulation-based reinforcement-learning approach to optimizing real-time content selection and digital inventory management.