How can deep neural networks fail even with global optima?

DOI10.4208/ijnam2024-1027MaRDI QIDQ6648519

Qingguang Guan

Publication date: 4 December 2024

Published in: International Journal of Numerical Analysis and Modeling (Search for Journal in Brave)

zbMATH Keywords

function approximationbinary classificationglobal optimaoverfittingdeep neural network

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Applications to the sciences (65Z05) Numerical approximation and computational geometry (primarily algorithms) (65D99)

Cites Work

• High-order approximation rates for shallow neural networks with cosine and \(\mathrm{ReLU}^k\) activation functions
• Uniform approximation rates and metric entropy of shallow neural networks
• Error bounds for approximations with deep ReLU networks
• Relu Deep Neural Networks and Linear Finite Elements
• Universal approximation bounds for superpositions of a sigmoidal function
• Approximation by superpositions of a sigmoidal function

This page was built for publication: How can deep neural networks fail even with global optima?