Research project
Scientific machine learning
Physics-aware forecasting and hierarchical experts for ocean carbon dynamics.
My master’s research at the University of Chinese Academy of Sciences, advised by Prof. Jungang Xu, includes learning systems for complex physical processes.
Carbon-ECS introduces a 17-year East China Sea benchmark and combines large-scale transport modeling with local detail refinement. REM-MoE extends this line of work through hierarchical experts that separate environmental regimes from physical mechanisms.
I am interested in representations that make long-horizon physical prediction more reliable and interpretable.
