[en] Several learning algorithms in classification and structured prediction are formulated as large scale optimization problems. We show that a generic iterative reformulation and resolving strategy based on the progressive hedging algorithm from stochastic programming results in a highly parallel algorithm when applied to the large margin classification problem with nonlinear kernels. We also underline promising aspects of the available analysis of progressive hedging strategies.
Research Center/Unit :
Systems and Modeling Research Unit
Disciplines :
Computer science
Author, co-author :
Defourny, Boris ; Université de Liège - ULiège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Dép. d'électric., électron. et informat. (Inst.Montefiore)
Wehenkel, Louis ; Université de Liège - ULiège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Systèmes et modélisation
Language :
English
Title :
Large Margin Classification with the Progressive Hedging Algorithm
Publication date :
December 2009
Event name :
Second NIPS Workshop on Optimization for Machine Learning
Event organizer :
Sebastian Nowozin; Suvrit Sra; SVN Vishwanathan; Stephen Wright
Event place :
Whistler, Canada
Event date :
December 12, 2009
Audience :
International
Funders :
DYSCO (Dynamical Systems, Control, and Optimization); PASCAL2 Network of Excellence