Non-asymptotic estimates for TUSLA algorithm for non-convex learning with applications to neural networks with ReLU activation function

DOI10.1093/imanum/drad038MaRDI QIDQ6631413

Ariel Neufeld, Dong-Young Lim, Ying Zhang, Sotirios Sabanis

Publication date: 1 November 2024

Published in: IMA Journal of Numerical Analysis (Search for Journal in Brave)

zbMATH Keywords

nonconvex optimizationartificial neural networksnonasymptotic estimatesReLU activation functionsuper-linearly growing coefficientsdiscontinuous stochastic gradienttaming technique

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07)

This page was built for publication: Non-asymptotic estimates for TUSLA algorithm for non-convex learning with applications to neural networks with ReLU activation function