{"id":"arxiv/2602.17776","url":"https://arxiv.org/abs/2602.17776","tags":["math.NA","cs.LG"],"score":0,"title":"Solving and learning advective multiscale Darcian dynamics with the Neural Basis Method","domain":"math.NA","source":"arxiv","authors":["Yuhe Wang","Min Wang"],"reviews":0,"subject":"math.NA","summary":"Physics-governed models are increasingly paired with machine learning for accelerated predictions, yet most \"physics--informed\" formulations treat the governing equations as a penalty loss whose scale and meaning are set by heuristic balancing. This blurs operator structure, thereby confounding solution approximation error with governing-equation enforcement error and making the solving and learning progress hard to interpret and control. Here we introduce the Neural Basis Method, a projection-based formulation that couples a predefined, physics-conforming neural basis space with an…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-02-19","reactions":{}}