000265561 001__ 265561
000265561 005__ 20190909131035.0
000265561 037__ $$aCONF
000265561 245__ $$aA Conditional Gradient-Based Augmented Lagrangian Framework
000265561 260__ $$c2019
000265561 269__ $$a2019
000265561 336__ $$aConference Papers
000265561 500__ $$aScheduled publication of Proceedings: Volume 97 is assigned to ICML 2019 (ISSN: 2640-3498)
000265561 520__ $$aThis paper considers a generic convex minimization template with affine constraints over a compact domain, which covers key semidefinite programming applications. The existing conditional gradient methods either do not apply to our template or are too slow in practice. To this end, we propose a new conditional gradient method, based on a unified treatment of smoothing and augmented Lagrangian frameworks. The proposed method maintains favorable properties of the classical conditional gradient method, such as cheap linear minimization oracle calls and sparse representation of the decision variable. We prove O(1/√k) convergence rate of our method in the objective residual and the feasibility gap. This rate is essentially the same as the state of the art CG-type methods for our problem template, but the proposed method is significantly superior to existing methods in various semidefinite programming applications.
000265561 6531_ $$aml-tm
000265561 6531_ $$aml-ai
000265561 700__ $$0248415$$aYurtsever, Alp$$g233086
000265561 700__ $$aFercoq, Olivier
000265561 700__ $$0243957$$aCevher, Volkan$$g199128
000265561 7112_ $$dJune 9-15, 2019$$cLong Beach, USA$$a36th International Conference on Machine Learning (ICML 2019)
000265561 773__ $$tProceedings of the International Conference on Machine Learning - ICML 2019
000265561 8560_ $$falp.yurtsever@epfl.ch
000265561 8564_ $$uhttps://infoscience.epfl.ch/record/265561/files/YFC2019.pdf$$zFinal$$s1291101
000265561 909C0 $$zMarselli, Béatrice$$xU12179$$pLIONS$$mvolkan.cevher@epfl.ch$$0252306
000265561 909CO $$pconf$$pSTI$$ooai:infoscience.epfl.ch:265561
000265561 960__ $$agosia.baltaian@epfl.ch
000265561 961__ $$aalessandra.bianchi@epfl.ch
000265561 973__ $$rREVIEWED$$aEPFL
000265561 980__ $$aCONF
000265561 981__ $$aoverwrite