On the Global Linear Convergence of Frank-Wolfe Optimization Variants

Simon Lacoste-Julien 1, 2, 3, * Martin Jaggi 4
* Corresponding author
1 SIERRA - Statistical Machine Learning and Parsimony
DI-ENS - Département d'informatique de l'École normale supérieure, ENS Paris - École normale supérieure - Paris, Inria Paris-Rocquencourt, CNRS - Centre National de la Recherche Scientifique : UMR8548
Abstract : The Frank-Wolfe (FW) optimization algorithm has lately re-gained popularity thanks in particular to its ability to nicely handle the structured constraints appearing in machine learning applications. However, its convergence rate is known to be slow (sublinear) when the solution lies at the boundary. A simple less-known fix is to add the possibility to take 'away steps' during optimization, an operation that importantly does not require a feasibility oracle. In this paper, we highlight and clarify several variants of the Frank-Wolfe optimization algorithm that have been successfully applied in practice: away-steps FW, pairwise FW, fully-corrective FW and Wolfe's minimum norm point algorithm, and prove for the first time that they all enjoy global linear convergence, under a weaker condition than strong convexity of the objective. The constant in the convergence rate has an elegant interpretation as the product of the (classical) condition number of the function with a novel geometric quantity that plays the role of a 'condition number' of the constraint set. We provide pointers to where these algorithms have made a difference in practice, in particular with the flow polytope, the marginal polytope and the base polytope for submodular optimization.
Complete list of metadatas

https://hal.inria.fr/hal-01248675
Contributor : Simon Lacoste-Julien <>
Submitted on : Monday, December 28, 2015 - 4:49:42 AM
Last modification on : Tuesday, March 5, 2019 - 9:30:10 AM

Links full text

Identifiers

  • HAL Id : hal-01248675, version 1
  • ARXIV : 1511.05932

Collections

Citation

Simon Lacoste-Julien, Martin Jaggi. On the Global Linear Convergence of Frank-Wolfe Optimization Variants. NIPS 2015 - Advances in Neural Information Processing Systems 28, Dec 2015, Montreal, Canada. ⟨hal-01248675⟩

Share

Metrics

Record views

267