Optimal learning of high-dimensional classification problems using deep neural networks

MaRDI QIDQ6386542zbMATH OpenFDO

Authors Philipp Petersen, Felix Voigtlaender

Publication date 23 December 2021

Mathematics Subject Classification ID

Rate of convergence, degree of approximation (41A25) Approximation by arbitrary nonlinear expressions; widths and entropy (41A46) Minimax procedures in statistical decision theory (62C20) Learning and adaptive systems in artificial intelligence (68T05)

Abstract: We study the problem of learning classification functions from noiseless training samples, under the assumption that the decision boundary is of a certain regularity. We establish universal lower bounds for this estimation problem, for general classes of continuous decision boundaries. For the class of locally Barron-regular decision boundaries, we find that the optimal estimation rates are essentially independent of the underlying dimension and can be realized by empirical risk minimization methods over a suitable class of deep neural networks. These results are based on novel estimates of the $L^1$ and $L^{i}nfty$ entropies of the class of Barron-regular functions.

This page was built for publication: Optimal learning of high-dimensional classification problems using deep neural networks

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6386542)