{"id":"arxiv/2607.06287","url":"https://arxiv.org/abs/2607.06287","tags":["math.NA","cs.LG"],"score":0,"title":"Kernel-based Operator Learning: Error Analysis, Budget Allocation, and a Physics-Informed Extension","domain":"math.NA","source":"arxiv","authors":["Rüdiger Kempf"],"reviews":0,"subject":"math.NA","summary":"We study kernel-based operator learning in a two-stage sampling framework, where an offline kernel regression operator learns a discretized representation of the target operator from input-output pairs and an online kernel reconstruction operator recovers the output function from predicted observations. Our main theoretical contribution is an explicit budget allocation condition relating the number $N$ of training pairs, the number $n$ of input observations, and the output resolution $m$. The condition is derived from a coupled error analysis that interprets the surrogate as a reconstruction…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-07-07","reactions":{}}