Summary¶
I am a computational physicist specializing in numerical and statistical methods, applied to high-resolution large-eddy simulation (LES) models of the atmosphere. I am particularly interested in implementing and optimizing parallel processing of large, multi-dimensional data, as well as machine learning algorithms such as Gaussian Process (GP) and deep neural networks to better represent the complex dynamics seen in the modelled data.
Skills¶
15+ years of project experience with
Python
,C
, andC++
Parallel processing of multi-dimensional data with
C++
(xtensor) andJulia
Statistical analysis with
Numpy
,Scipy
andScikit-learn
Data pre-processing with
Pandas
/Polars
andArrow/Parquet
Machine learning experience with
Pytorch
[1],Keras
andJax
Experience with RDBMS with
MongoDB
,PostgresSQL
andMariaDB
Education¶
Ph.D. in Atmospheric Science | 2014 - 2024 |
---|---|
University of British Columbia Vancouver, BC, Canada |
M.Sc. in Atmospheric Science | 2012 - 2014 |
---|---|
McGill University Montreal, Quebec, Canada |
B.Sc. Combined Honours in Computer Science and Physics | 2008 - 2012 |
---|---|
University of British Columbia Vancouver, BC, Canada |
More information is available here.
Employment¶
Research Scientist | 2020 - 2023 |
---|---|
Korea Polar Research Institute (KOPRI) Incheon, Korea |
More information is available here.
Publication¶
See the full list of publications.
See this Jupyter Book, which is a collection of Jupyter notebooks accompanying Oh and Austin, 2015.
- Oh, G., & Austin, P. H. (2024). Quantifying the Oscillatory Evolution of Simulated Boundary-Layer Cloud Fields Using Gaussian Process Regression. 10.5194/egusphere-2024-352