Projects

Physics-based Machine Learning Scheme for Accelerating Simulation of 2D Lower-Hybrid Wave on TST-2 Spherical Tokamak

Physics-based Machine Learning Scheme for Accelerating Simulation of 2D Lower-Hybrid Wave on TST-2 Spherical Tokamak

Physics-informed neural operator surrogate for lower hybrid wave simulation on the TST-2 spherical tokamak, combining plasma physics, scientific machine learning, and fusion-oriented HPC.
Pretrained Fourier Neural Operator for Non-Newtonian Fluid Dynamics

Pretrained Fourier Neural Operator for Non-Newtonian Fluid Dynamics

Fourier Neural Operator surrogate modeling for non-Newtonian fluid dynamics in Moldex3D, focused on scientific computing and operator learning for PDEs.
Regression-based Physics Informed Neural Networks for Magnetopause Tracking

Regression-based Physics Informed Neural Networks for Magnetopause Tracking

Reg-PINNs model for magnetopause tracking that combines physics-based regression and neural networks to improve scientific machine learning generalization.
Parallel Computing, with applications in Quantitative Strategies

Parallel Computing, with applications in Quantitative Strategies

Parallel computing project benchmarking multicore quantitative strategies on an HPC cluster.
Numerical Methods and Physics Informed Neural Networks in Advection Function

Numerical Methods and Physics Informed Neural Networks in Advection Function

PINN and numerical-method study of the advection equation, comparing discretization strategies and physics-informed learning for PDEs.
Comparing Kalman Filter with Recurrent-based machine learning on Satellite Attitude Determination

Comparing Kalman Filter with Recurrent-based machine learning on Satellite Attitude Determination

Kozak Hou, Jih-Hong Shue
Automatic Emergency Dust-Free solution on-board International Space Station with Bi-GRU

Automatic Emergency Dust-Free solution on-board International Space Station with Bi-GRU

Kozak Hou, Wei-Chih Lin, Hong-Chun Hou, Yu-Hao Huang, Jih-Hong Shue
High Performance Computing with NVIDIA-RAPIDS

High Performance Computing with NVIDIA-RAPIDS

High-performance computing project using NVIDIA RAPIDS for GPU-accelerated data science workflows on EMNIST and remote sensing datasets.