A generic approach for escaping saddle points

Sashank J. Reddi, Manzil Zaheer, Suvrit Sra, Barnabás Póczos, Ruslan Salakhutdinov, Francis Bach, Alexander J. Smola

Research output: Contribution to conferencePaperpeer-review

33 Scopus citations

Abstract

A central challenge to using first-order methods for optimizing nonconvex problems is the presence of saddle points. First-order methods often get stuck at saddle points, greatly deteriorating their performance. Typically, to escape from saddles one has to use second-order methods. However, most works on second-order methods rely extensively on expensive Hessian-based computations, making them impractical in large-scale settings. To tackle this challenge, we introduce a generic framework that minimizes Hessian-based computations while at the same time provably converging to second- order critical points. Our framework carefully alternates between a first-order and a second-order subroutine, using the latter only close to saddle points, and yields convergence results competitive to the state-of-the-art. Empirical results suggest that our strategy also enjoys a good practical performance.

Original languageEnglish
Pages1233-1242
Number of pages10
StatePublished - 2018
Externally publishedYes
Event21st International Conference on Artificial Intelligence and Statistics, AISTATS 2018 - Playa Blanca, Lanzarote, Canary Islands, Spain
Duration: 9 Apr 201811 Apr 2018

Conference

Conference21st International Conference on Artificial Intelligence and Statistics, AISTATS 2018
Country/TerritorySpain
CityPlaya Blanca, Lanzarote, Canary Islands
Period9/04/1811/04/18

Fingerprint

Dive into the research topics of 'A generic approach for escaping saddle points'. Together they form a unique fingerprint.

Cite this