DocumentCode :
3537938
Title :
Convergence analysis of primal solutions in dual first-order methods
Author :
Jie Lu ; Johansson, Mikael
Author_Institution :
ACCESS Linnaeus Center, KTH R. Inst. of Technol., Stockholm, Sweden
fYear :
2013
fDate :
10-13 Dec. 2013
Firstpage :
6861
Lastpage :
6867
Abstract :
This paper studies primal convergence in dual first-order methods for convex optimization. Specifically, we consider Lagrange decomposition of a general class of inequality- and equality-constrained optimization problems with strongly convex, but not necessarily differentiable, objective functions. The corresponding dual problem is solved using a first-order method, and the minimizer of the Lagrangian computed when evaluating the dual function is considered as an approximate primal solution. We derive error bounds for this approximate primal solution in terms of the dual errors. Based on such error bounds, we show that the approximate primal solution converges to the primal optimum at a rate no worse than O(1/√k) if the projected dual gradient method is adopted and O(1/k) if a fast gradient method is utilized, where k is the number of iterations. Finally, via simulation, we compare the convergence behavior of different approximate primal solutions in various dual first-order methods in the literature.
Keywords :
approximation theory; constraint theory; convergence; convex programming; duality (mathematics); gradient methods; Lagrange decomposition; approximate primal solution; convergence behavior; convex optimization; dual errors; dual first-order methods; dual function; dual gradient method; equality-constrained optimization problems; error bounds; inequality; iterations; primal convergence analysis; primal solutions; Convergence; Convex functions; Euclidean distance; Gradient methods; Linear programming; Upper bound;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Decision and Control (CDC), 2013 IEEE 52nd Annual Conference on
Conference_Location :
Firenze
ISSN :
0743-1546
Print_ISBN :
978-1-4673-5714-2
Type :
conf
DOI :
10.1109/CDC.2013.6760976
Filename :
6760976
Link To Document :
بازگشت