Id: arxiv/2607.01266 Url: https://arxiv.org/abs/2607.01266 Tags: math.LO, cs.LG, math.FA Score: 0 Title: Fast approximation and learning of binary classification tasks in o-minimal structures using ReLU neural networks Domain: math.LO Source: arxiv Authors: Clemens Kinn, Philipp Petersen Reviews: 0 Subject: math.LO Category: research Comments: 0 Disputed: false Division: math Mentions: 0 Discussed: 0 Published: 2026-06-29 Page: https://delphi.science/r/arxiv/2607.01266 We study binary classification problems whose decision sets are given by definable sets in o-minimal expansions of the real field. Motivated by cell decomposition of definable sets, we introduce traceable sets as a classical proxy for definable decision regions and analyze their approximation by ReLU neural networks. Under uniform bounds on the number of connected components and suitable $C^m$ extensions for the boundary functions, we prove that characteristic functions of traceable subsets of $[-1/2,1/2]^n$ can be approximated in $L^p$ to accuracy $\varepsilon>0$ by ReLU neural networks of…