- USA
- Proposed a neural dense regression model that maps varying coefficient fields to solutions (a mapping between two function spaces) in elliptic PDEs
- Our network beat a reduced-order projection method and other existing operators in accuracy by
- 30%, and showed greater data efficiency. The advantage was pronounced with high-frequency, highlycontrasted
M
Microsoft Research Asia
Research Assistant
Beijing, CN
April 2022 - September 2022
T
Tsinghua Laboratory of Brain and Intelligence
Research Assistant
Beijing, CN
June 2021 - April 2022
C
Citadel AND Citadel Securities Asia Pacific Datathon
March 2021 - March 2021
Skills
AlgorithmsArtificial Neural NetworksEuclidean GeometryFinancial SecuritiesForecasting SkillsGeometryKnowledge of EngineeringKnowledge of MathematicsPartial Differential EquationPhysicsResearch SkillsSoccerStrategic Thinking