Wenxia Wei

National University of Defense Technology

Papers

Total Citations

H-Index

About

Wenxia Wei is a researcher whose work bridges the gap between data efficiency and scalability in reinforcement learning, with a particular focus on robotic control. Her key research areas include model-free and model-based reinforcement learning, high-dimensional action spaces, and efficient learning algorithms for autonomous systems. Wei's major contribution lies in developing methods that optimize high-dimensional learners by leveraging low-dimensional action features, enabling robots to learn complex tasks more efficiently without requiring massive training datasets. Her 2019 paper, "Optimizing High-dimensional Learner with Low-Dimension Action Features," has garnered 4 citations and addresses a critical bottleneck in robotics: the trade-off between the scalability of model-free methods and the sample efficiency of model-based approaches. By proposing a framework that extracts low-dimensional action representations, Wei demonstrates how to reduce training time while maintaining performance on high-dimensional tasks. Her work is particularly notable for its practical implications, offering a pathway toward more data-efficient robotic learning in real-world applications. Wei's research continues to influence the field of reinforcement learning, inspiring new approaches to handling the curse of dimensionality in autonomous systems.