AI RESEARCH
Generating Rectifiable Measures through Neural Networks
arXiv CS.LG
•
ArXi:2412.05109v2 Announce Type: replace We derive universal approximation results for the class of (countably) $m$-rectifiable measures. Specifically, we prove that $m$-rectifiable measures can be approximated as push-forwards of the one-dimensional Lebesgue measure on $[0,1]$ using ReLU neural networks with arbitrarily small approximation error in terms of Wasserstein distance.