Muhammad Akbar Khan

MS Applied Mathematics researcher at NED University of Engineering & Technology, Karachi. I develop physics-informed neural networks for solving partial differential equations, with a focus on interface-tracking problems in computational fluid dynamics.

Currently seeking fully-funded PhD positions in scientific machine learning, neural operators, and data-driven PDE modeling.

Research Interests

Physics-Informed Neural Networks

Training strategies, loss balancing, causal weighting, and architectures (RFF, modified MLPs) for accurate PDE solutions.

Level-Set Methods

Neural network approaches to interface tracking and advection in computational fluid dynamics.

Neural Operators

Foundation models and operator learning (DeepONet, FNO) for cross-domain PDE solving and surrogate modeling.

Scientific Computing

Numerical methods for PDEs, optimization, and high-performance computing for scientific applications.

Publications

Under Review

A Systematic Study of Physics-Informed Neural Networks for Level-Set Interface Advection

Muhammad Akbar Khan and Fahim Raees

International Journal for Numerical Methods in Fluids (IJNMF), Wiley, 2026

A comprehensive study of PINNs for solving the level-set advection equation across four benchmark problems: translation, rigid-body rotation, Zalesak's disk, and the reversed single-vortex. 47 experiments systematically investigate network architecture (Tanh vs. RFF), training strategies (causal weighting, learning rate scheduling), eikonal regularization, and collocation sampling. The proposed PINN framework achieves results approximately 2× more accurate than the state-of-the-art classical numerical method on the reversed vortex benchmark.

Education

2023 — 2026

MS Applied Mathematics

NED University of Engineering & Technology, Karachi, Pakistan

Thesis: Physics-Informed Neural Networks for the Level-Set Equation

Supervisor: Dr. Fahim Raees

Technical Skills

Python PyTorch NumPy Matplotlib SciPy Pandas LaTeX Google Colab Git/GitHub PostgreSQL Jupyter

Contact

I am actively seeking fully-funded PhD positions (Fall 2026/2027) in scientific machine learning, with a focus on physics-informed methods and neural operators for PDE modeling. If you are interested in my work or have opportunities, I would be glad to hear from you.

Email: khan.pg4200844@cloud.neduet.edu.pk
ORCID: 0009-0001-7956-0080
GitHub: AkbarTheAnalyst