L-MAU-Cahn-Hilliard: MPI Solver and Low-Dimensional Prediction Pipeline
Published:
Status: Published
Project Goal:
Build an end-to-end workflow for Cahn-Hilliard microstructure evolution, including MPI-based simulation, dimensionality reduction (PCA/autoencoder pipelines), and sequence prediction with an L-MAU model.
My Role:
Developed and maintained the simulation and ML workflow, integrated data pipelines, and organized training/testing steps for reduced-order prediction.
Tech Stack:
C, MPI, FFTW3, GSL, Python, NumPy, PCA, Autoencoder, L-MAU
Repository:
https://github.com/ShengJer/L-MAU-Cahn-Hilliard