Id: arxiv/2603.06380 Url: https://arxiv.org/abs/2603.06380 Tags: math.NA, cs.AI, cs.LG Score: 0 Title: Kinetic-based regularization: Learning spatial derivatives and PDE applications Domain: math.NA Source: arxiv Authors: Abhisek Ganguly, Santosh Ansumali, Sauro Succi Reviews: 0 Subject: math.NA Category: research Comments: 0 Disputed: false Division: math Mentions: 0 Discussed: 0 Published: 2026-03-06 Page: https://delphi.science/r/arxiv/2603.06380 Accurate estimation of spatial derivatives from discrete and noisy data is central to scientific machine learning and numerical solutions of PDEs. We extend kinetic-based regularization (KBR), a localized multidimensional kernel regression method with a single trainable parameter, to learn spatial derivatives with provable second-order accuracy in 1D. Two derivative-learning schemes are proposed: an explicit scheme based on the closed-form prediction expressions, and an implicit scheme that solves a perturbed linear system at the points of interest. The fully localized formulation enables…