Skip to main content
Cornell University
We gratefully acknowledge support from the Simons Foundation, member institutions, and all contributors. Donate
arxiv logo > cs > arXiv:2002.06189

Help | Advanced Search

arXiv logo
Cornell University Logo

quick links

  • Login
  • Help Pages
  • About

Computer Science > Machine Learning

arXiv:2002.06189 (cs)
[Submitted on 14 Feb 2020 (v1), last revised 2 Nov 2020 (this version, v2)]

Title:Stochasticity of Deterministic Gradient Descent: Large Learning Rate for Multiscale Objective Function

Authors:Lingkai Kong, Molei Tao
View a PDF of the paper titled Stochasticity of Deterministic Gradient Descent: Large Learning Rate for Multiscale Objective Function, by Lingkai Kong and Molei Tao
View PDF
Abstract:This article suggests that deterministic Gradient Descent, which does not use any stochastic gradient approximation, can still exhibit stochastic behaviors. In particular, it shows that if the objective function exhibit multiscale behaviors, then in a large learning rate regime which only resolves the macroscopic but not the microscopic details of the objective, the deterministic GD dynamics can become chaotic and convergent not to a local minimizer but to a statistical distribution. A sufficient condition is also established for approximating this long-time statistical limit by a rescaled Gibbs distribution. Both theoretical and numerical demonstrations are provided, and the theoretical part relies on the construction of a stochastic map that uses bounded noise (as opposed to discretized diffusions).
Comments: NeurIPS 2020. v1->v2: Weakened conditions needed for the theory. Added connections to neural network. Corrected typo
Subjects: Machine Learning (cs.LG); Dynamical Systems (math.DS); Numerical Analysis (math.NA); Optimization and Control (math.OC); Machine Learning (stat.ML)
Cite as: arXiv:2002.06189 [cs.LG]
  (or arXiv:2002.06189v2 [cs.LG] for this version)
  https://doi.org/10.48550/arXiv.2002.06189
arXiv-issued DOI via DataCite

Submission history

From: Lingkai Kong [view email]
[v1] Fri, 14 Feb 2020 18:59:20 UTC (2,322 KB)
[v2] Mon, 2 Nov 2020 16:37:14 UTC (2,557 KB)
Full-text links:

Access Paper:

    View a PDF of the paper titled Stochasticity of Deterministic Gradient Descent: Large Learning Rate for Multiscale Objective Function, by Lingkai Kong and Molei Tao
  • View PDF
  • TeX Source
  • Other Formats
view license
Current browse context:
cs.LG
< prev   |   next >
new | recent | 2020-02
Change to browse by:
cs
cs.NA
math
math.DS
math.NA
math.OC
stat
stat.ML

References & Citations

  • NASA ADS
  • Google Scholar
  • Semantic Scholar

DBLP - CS Bibliography

listing | bibtex
Lingkai Kong
Molei Tao
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