PINNs

An inverse physics-informed neural network that recovers a material's thermal diffusivity from sparse, noisy temperature measurements, benchmarked against classical parameter fitting.

A self-directed curriculum working from mathematical foundations toward physics-informed machine learning, implemented from scratch.