Aeronautical Engineering student at Imperial College London, currently on a year in industry at Siemens DISW, working on CFD software and numerical tools.

• Neural networks implemented from first principles (autograd, backpropagation, optimisation)
• Function approximation and generalisation behaviour
• Comparison of classical numerical methods and ML surrogates

• approx_Sin — neural network approximation of sin(πx) using scratch and PyTorch implementations (Completed)
• HKJC_ML — machine learning pipelines for noisy real-world horse racing data from the Hong Kong Jockey Club (Private)
• nn-aerofoil-surrogate-distribution-shift — neural-network surrogate modelling for aerofoil aerodynamic characteristics (Completed)
• pinn-pdes — follow-on project exploring physics-informed neural networks for PDEs, including heat, diffusion, and wave equations, after initial work on ML surrogates for aerodynamic prediction (In progress)