Home

Welcome! My name is Linfeng Zhao (赵林风), a first-year Ph.D. student working with Prof. Lawson Wong at Khoury College of Northeastern University. Feel free to contact me on related research or problems.

Biography

Linfeng Zhao is a first-year CS Ph.D. student advised by Prof. Lawson L.S. Wong at Khoury College of Computer Sciences of Northeastern University. His research interests lie in machine learning, robotics, and artificial intelligence, and is also motivated by computer vision and natural language.

A general problem Linfeng cares about is learning general and abstracted knowledge from the world, as well as representing and applying the learned knowledge in simulated or real-world environments. Specifically, he is currently focusing on learning abstracted state representation and structured dynamics model in reinforcement learning for various applications.

During his undergraduate studies, he interned at Microsoft Research Asia on reinforcement learning. Before that, he spent the defining time at UC San Diego and worked with Prof. Hao Su on machine learning and vision. He spent the first two years at Northeastern University in China and worked with Prof. Yuan Liu for his B.E. in pilot class in software engineering.

He enjoys sports, gaming, travelling, photography, etc.

Selected Publications

  • Parameterized Action Reinforcement Learning for Inverted Index Match Plan Generation
    • Linfeng Zhao, …
    • In submission towards ML conference
  • InterFact: Towards Interactive Factorization of Actionable Entities
    • Linfeng Zhao, Lawson L.S. Wong, Hao Su
    • In active progress

Research Experience

  • PhD Research Assistant @ Northeastern University advised by Prof. Lawson Wong
    • Projects in progress.
  • Research Intern @ Microsoft Research Asia with Dr. Qi Chen
    • Parameterized Action Reinforcement Learning for Inverted Index Match Plan Generation
  • Undergraduate Research Assistant @ UC San Diego advised by Prof. Hao Su (SU Lab)
    • InterFact: Towards Interactive Factorization of Actionable Entities
    • PartNet: A Large-scale Benchmark for Fine-grained and Hierarchical Part-level 3D Object Understanding