Education
National University of Singapore
Degree: Bachelor of Science (Honours, Highest Distinction) in Physics
Completion Date: 30 June 2020
Thesis: Density Potential Functional Theory in position and momentum space and its implementation using the Machine Learning library PyTorch
- Summary: This thesis explores Density Potential Functional Theory (DPFT), introduced by Julian Schwinger and Berge Englert, focusing on its application to a spin-polarized Fermi gas with magnetic dipole-dipole interaction under the Thomas Fermi (TF) approximation in both position and momentum spaces. A key contribution is the development of an original orbital-free DPFT code, leveraging the PyTorch machine learning library for multi-GPU acceleration, and its release as a Python package. The code accurately implements 2D and 3D magnetic dipole-dipole interactions, with results aligning with physical expectations. Comparisons with custom Kohn-Sham DFT code and the state-of-the-art software VASP were conducted to enhance accuracy. An innovative low-pass filter on TF density improved results, suggesting future research potential.
Verify the certificate at https://www.opencerts.io
Thesis
Study Notes (BSc)
