Repository logo

Infoscience

  • English
  • French
Log In
Logo EPFL, École polytechnique fédérale de Lausanne

Infoscience

  • English
  • French
Log In
  1. Home
  2. Academic and Research Output
  3. Conferences, Workshops, Symposiums, and Seminars
  4. Generalization of Quasi-Newton Methods: Application to Robust Symmetric Multisecant Updates
 
conference paper

Generalization of Quasi-Newton Methods: Application to Robust Symmetric Multisecant Updates

Scieur, Damien
•
Liu, Lewis
•
Pumir, Thomas
Show more
January 1, 2021
24Th International Conference On Artificial Intelligence And Statistics (Aistats)
24th International Conference on Artificial Intelligence and Statistics (AISTATS)

Quasi-Newton (qN) techniques approximate the Newton step by estimating the Hessian using the so-called secant equations. Some of these methods compute the Hessian using several secant equations but produce non-symmetric updates. Other quasi-Newton schemes, such as BFGS, enforce symmetry but cannot satisfy more than one secant equation. We propose a new type of quasi-Newton symmetric update using several secant equations in a least-squares sense. Our approach generalizes and unifies the design of quasi-Newton updates and satisfies provable robustness guarantees.

  • Details
  • Metrics
Type
conference paper
Web of Science ID

WOS:000659893800062

Author(s)
Scieur, Damien
Liu, Lewis
Pumir, Thomas
Boumal, Nicolas  
Date Issued

2021-01-01

Publisher

MICROTOME PUBLISHING

Publisher place

Brookline

Published in
24Th International Conference On Artificial Intelligence And Statistics (Aistats)
Series title/Series vol.

Proceedings of Machine Learning Research; 130

Start page

550

End page

558

Subjects

Computer Science, Artificial Intelligence

•

Mathematics, Applied

•

Statistics & Probability

•

Computer Science

•

Mathematics

•

convergence

•

bfgs

Editorial or Peer reviewed

REVIEWED

Written at

EPFL

EPFL units
OPTIM  
Event nameEvent placeEvent date
24th International Conference on Artificial Intelligence and Statistics (AISTATS)

ELECTR NETWORK

Apr 13-15, 2021

Available on Infoscience
August 28, 2021
Use this identifier to reference this record
https://infoscience.epfl.ch/handle/20.500.14299/180985
Logo EPFL, École polytechnique fédérale de Lausanne
  • Contact
  • infoscience@epfl.ch

  • Follow us on Facebook
  • Follow us on Instagram
  • Follow us on LinkedIn
  • Follow us on X
  • Follow us on Youtube
AccessibilityLegal noticePrivacy policyCookie settingsEnd User AgreementGet helpFeedback

Infoscience is a service managed and provided by the Library and IT Services of EPFL. © EPFL, tous droits réservés