Provably training overparameterized neural network classifiers with non-convex constraints

DOI10.1214/22-EJS2036MaRDI QIDQ2106783

Mladen Kolar, You-Lin Chen, Zhaoran Wang

Publication date: 19 December 2022

Published in: Electronic Journal of Statistics (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/2012.15274

zbMATH Keywords

neural tangent kernelnon-convex constrained optimizationonline learning with non-convex losses

Mathematics Subject Classification ID

Statistics (62-XX)

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Corrigendum to: ``On the complexity of finding first-order critical points in constrained nonlinear optimization
• A selective overview of deep learning
• Gradient descent optimizes over-parameterized deep ReLU networks
• Stochastic first-order methods for convex and nonconvex functional constrained optimization
• Fairness through awareness
• A Dynamic Near-Optimal Algorithm for Online Linear Programming
• Online Learning and Online Convex Optimization
• Stochastic Model-Based Minimization of Weakly Convex Functions
• Non-convex Optimization for Machine Learning
• Agnostic Learning of Monomials by Halfspaces Is Hard
• Online Linear Programming: Dual Convergence, New Algorithms, and Regret Bounds
• Tensor Canonical Correlation Analysis With Convergence and Statistical Guarantees
• Optimization with Non-Differentiable Constraints with Applications to Fairness, Recall, Churn, and Other Goals
• Wide neural networks of any depth evolve as linear models under gradient descent ^*
• Advancing subgroup fairness via sleeping experts
• An adaptive stochastic sequential quadratic programming with differentiable exact augmented Lagrangians

This page was built for publication: Provably training overparameterized neural network classifiers with non-convex constraints