Skip to content

Questions I'm working on

Research.

My research spans scientific computing, machine learning and mathematical modeling. Choose a topic below to read about the question, the approach and the work so far.

Current research

Hidden source
The original letter-A heat source from the SRF demo.
Measured heat
Temperature measurements blur the source into a diffuse heat field.
Recovered shape
The letter-trained POD model recovers a recognizable A outline.
Source → measured temperature → recovered outline · SRF demo

HKU Summer Research Fellowship · 2026

Recovering a hidden heat source

Heat diffusion blurs a source’s shape. I study how to recover useful shape information from temperature measurements, and when a smaller model can make the computation faster. I was selected as one of the best poster presenters for this work.

Inverse problems · Scientific computing · POD

Watch the demo & read the research
Illustrative SGD and SAM solutions on an uneven loss curve: a sharp minimum versus a broad basin, with equal-size parameter neighborhoods.
Sharp point, broad basin · Illustrative SGD/SAM comparison

Current capstone · In progress

Sharpness-aware optimization

I’m exploring sharpness-aware optimization for machine-learning training. This page will follow the capstone as the experiments and findings take shape.

Machine learning · Optimization

View the current project

Previous research

Color-coded Hong Kong MTR route map showing lines and interchange stations.
Hong Kong rail network · Illustrative context · Emphrase, CC0

Independent research · 2023

Urban rail network optimization

How can a rail route balance geometry, engineering constraints and construction cost? I combined analytical optimization with Particle Swarm Optimization to study feasible routes.

Mathematical modeling · Route design

Read the project overview
NASA artist’s concept of the Milky Way, with spiral arms wrapping around a bright central bar.
Milky Way artist’s concept · NASA/JPL-Caltech

Earlier astronomy work

Mapping Galactic hydrogen

I used 21 cm neutral-hydrogen observations and computational models of Galactic rotation to study the large-scale structure of the Milky Way.

Astronomy · Computational modeling

Read the project overview
← Back to the homepageGet in touch ↗