{"id":"arxiv/2603.26418","url":"https://arxiv.org/abs/2603.26418","tags":["stat.ML","cs.LG","math.FA"],"score":0,"title":"Kantorovich--Kernel Neural Operators: Approximation Theory, Asymptotics, and Neural Network Interpretation","domain":"stat.ML","source":"arxiv","authors":["Tian-Xiao He"],"reviews":0,"subject":"stat.ML","summary":"This paper studies a class of multivariate Kantorovich-kernel neural network operators, including the deep Kantorovich-type neural network operators studied by Sharma and Singh. We prove density results, establish quantitative convergence estimates, derive Voronovskaya-type theorems, analyze the limits of partial differential equations for deep composite operators, prove Korovkin-type theorems, and propose inversion theorems. This paper studies a class of multivariate Kantorovich-kernel neural network operators, including the deep Kantorovich-type neural network operators studied by Sharma…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-03-27","reactions":{}}