{"id":"arxiv/2601.02492","url":"https://arxiv.org/abs/2601.02492","tags":["math.NA","cs.LG"],"score":0,"title":"Variational (Energy-Based) Spectral Learning: A Machine Learning Framework for Solving Partial Differential Equations","domain":"math.NA","source":"arxiv","authors":["M. M. Hammad"],"reviews":0,"subject":"math.NA","summary":"We introduce variational spectral learning (VSL), a machine learning framework for solving partial differential equations (PDEs) that operates directly in the coefficient space of spectral expansions. VSL offers a principled bridge between variational PDE theory, spectral discretization, and contemporary machine learning practice. The core idea is to recast a given PDE \\[ \\mathcal{L}u = f \\quad \\text{in} \\quad Q=Ω\\times(0,T), \\] together with boundary and initial conditions, into differentiable space-time energies built from strong-form least-squares residuals and weak (Galerkin)…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-01-05","reactions":{}}