Critical point-finding methods reveal gradient-flat regions of deep network losses
From MaRDI portal
Publication:5004367
DOI10.1162/NECO_A_01388OpenAlexW3159746799MaRDI QIDQ5004367FDOQ5004367
Authors: Charles G. Frye, James Simon, Neha S. Wadia, Andrew Ligeralde, Michael R DeWeese, Kristofer E. Bouchard 
Publication date: 30 July 2021
Published in: Neural Computation (Search for Journal in Brave)
Full work available at URL: https://arxiv.org/abs/2003.10397
Recommendations
- Shaping the learning landscape in neural networks around wide flat minima
- Landscape analysis for shallow neural networks: complete classification of critical points for affine target functions
- Optimization Landscape of Neural Networks
- The global optimization geometry of shallow linear neural networks
- Symmetry \& critical points for a model shallow neural network
Cites Work
- MINRES-QLP: a Krylov subspace method for indefinite or singular symmetric systems
- Title not available (Why is that?)
- Title not available (Why is that?)
- Title not available (Why is that?)
- Title not available (Why is that?)
- Numerically solving polynomial systems with Bertini
- Newton-Type Methods for Optimization and Variational Problems
- On the convergence of Newton iterations to non-stationary points
- Title not available (Why is that?)
- Analysis of Newton’s Method at Irregular Singularities
- The Fundamental Theorem of Linear Algebra
- Title not available (Why is that?)
- Implementable tensor methods in unconstrained convex optimization
- Residual and backward error bounds in minimum residual Krylov subspace methods
Cited In (1)
Uses Software
This page was built for publication: Critical point-finding methods reveal gradient-flat regions of deep network losses
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5004367)
