Skip to main content
Cornell University
We gratefully acknowledge support from the Simons Foundation, member institutions, and all contributors. Donate
arxiv logo > cs > arXiv:2102.02410

Help | Advanced Search

arXiv logo
Cornell University Logo

quick links

  • Login
  • Help Pages
  • About

Computer Science > Machine Learning

arXiv:2102.02410 (cs)
[Submitted on 4 Feb 2021 (v1), last revised 2 Jul 2021 (this version, v2)]

Title:A Local Convergence Theory for Mildly Over-Parameterized Two-Layer Neural Network

Authors:Mo Zhou, Rong Ge, Chi Jin
View a PDF of the paper titled A Local Convergence Theory for Mildly Over-Parameterized Two-Layer Neural Network, by Mo Zhou and 2 other authors
View PDF
Abstract:While over-parameterization is widely believed to be crucial for the success of optimization for the neural networks, most existing theories on over-parameterization do not fully explain the reason -- they either work in the Neural Tangent Kernel regime where neurons don't move much, or require an enormous number of neurons. In practice, when the data is generated using a teacher neural network, even mildly over-parameterized neural networks can achieve 0 loss and recover the directions of teacher neurons. In this paper we develop a local convergence theory for mildly over-parameterized two-layer neural net. We show that as long as the loss is already lower than a threshold (polynomial in relevant parameters), all student neurons in an over-parameterized two-layer neural network will converge to one of teacher neurons, and the loss will go to 0. Our result holds for any number of student neurons as long as it is at least as large as the number of teacher neurons, and our convergence rate is independent of the number of student neurons. A key component of our analysis is the new characterization of local optimization landscape -- we show the gradient satisfies a special case of Lojasiewicz property which is different from local strong convexity or PL conditions used in previous work.
Comments: Added references and fixed typos. Accepted to COLT 2021
Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML)
Cite as: arXiv:2102.02410 [cs.LG]
  (or arXiv:2102.02410v2 [cs.LG] for this version)
  https://doi.org/10.48550/arXiv.2102.02410
arXiv-issued DOI via DataCite

Submission history

From: Mo Zhou [view email]
[v1] Thu, 4 Feb 2021 04:41:04 UTC (96 KB)
[v2] Fri, 2 Jul 2021 18:13:24 UTC (96 KB)
Full-text links:

Access Paper:

    View a PDF of the paper titled A Local Convergence Theory for Mildly Over-Parameterized Two-Layer Neural Network, by Mo Zhou and 2 other authors
  • View PDF
  • TeX Source
  • Other Formats
view license
Current browse context:
cs.LG
< prev   |   next >
new | recent | 2021-02
Change to browse by:
cs
stat
stat.ML

References & Citations

  • NASA ADS
  • Google Scholar
  • Semantic Scholar

DBLP - CS Bibliography

listing | bibtex
Rong Ge
Chi Jin
a export BibTeX citation Loading...

BibTeX formatted citation

×
Data provided by:

Bookmark

BibSonomy logo Reddit logo

Bibliographic and Citation Tools

Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)

Code, Data and Media Associated with this Article

alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)

Demos

Replicate (What is Replicate?)
Hugging Face Spaces (What is Spaces?)
TXYZ.AI (What is TXYZ.AI?)

Recommenders and Search Tools

Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
IArxiv Recommender (What is IArxiv?)
  • Author
  • Venue
  • Institution
  • Topic

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)
  • About
  • Help
  • contact arXivClick here to contact arXiv Contact
  • subscribe to arXiv mailingsClick here to subscribe Subscribe
  • Copyright
  • Privacy Policy
  • Web Accessibility Assistance
  • arXiv Operational Status
    Get status notifications via email or slack