Papers
3
Total Citations
34
H-Index
3
About
Huixin Zhan is a researcher specializing in reinforcement learning, human-robot interaction, and multi-objective optimization, with a focus on developing intelligent systems that can learn complex behaviors in real-world environments. Her most recognized contribution, "Human-Guided Robot Behavior Learning: A GAN-Assisted Preference-Based Reinforcement Learning Approach," has garnered 27 citations and addresses a critical challenge in robotics: the difficulty of obtaining sufficient human demonstrations for training. By ingeniously integrating Generative Adversarial Networks (GANs) with preference-based reinforcement learning, Zhan's approach enables robots to learn nuanced behaviors even when direct human demonstration is impractical or insufficient. This work represents a meaningful bridge between human intuition and machine learning efficiency. Her earlier research on relationship-explainable multi-objective reinforcement learning further demonstrates her commitment to making AI systems not only performant but also interpretable, tackling the inherent tension between competing optimization objectives while generating semantic explanations for decision-making. Across her body of work, Zhan consistently pushes toward more trustworthy, human-aligned AI systems — a contribution of growing importance as autonomous robots and intelligent agents are increasingly deployed in complex, real-world settings.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3