Composite proximal bundle method

被引:29
作者
Sagastizabal, Claudia [1 ]
机构
[1] IMPA, BR-22460320 Rio De Janeiro, RJ, Brazil
关键词
Nonconvex Optimization; Nonsmooth optimization; Bundle methods; Composite functions; GAUSS-NEWTON METHOD; CONVEX-FUNCTIONS; CONVERGENCE; ALGORITHMS; SMOOTH;
D O I
10.1007/s10107-012-0600-5
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
We consider minimization of nonsmooth functions which can be represented as the composition of a positively homogeneous convex function and a smooth mapping. This is a sufficiently rich class that includes max-functions, largest eigenvalue functions, and norm-1 regularized functions. The bundle method uses an oracle that is able to compute separately the function and subgradient information for the convex function, and the function and derivatives for the smooth mapping. With this information, it is possible to solve approximately certain proximal linearized subproblems in which the smooth mapping is replaced by its Taylor-series linearization around the current serious step. Our numerical results show the good performance of the Composite Bundle method for a large class of problems.
引用
收藏
页码:189 / 233
页数:45
相关论文
共 34 条
[1]  
[Anonymous], 1993, Convex Analysis and Minimization Algorithms
[2]  
[Anonymous], PROXIMAL METHOD COMP
[3]  
[Anonymous], 1987, Unconstrained Optimization: Practical Methods of Optimization
[4]  
Bonnans J.F., 2000, SPRING S OPERAT RES, V1, DOI 10.1007/978-1-4612-1394-9
[5]  
Bonnans JF, 2006, NUMERICAL OPTIMIZATI, V2nd, DOI 10.1007/978-3-662-05078-1
[6]   A Gauss-Newton method for convex composite optimization [J].
Burke, JV ;
Ferris, MC .
MATHEMATICAL PROGRAMMING, 1995, 71 (02) :179-194
[7]   Two numerical methods for optimizing matrix stability [J].
Burke, JV ;
Lewis, AS ;
Lewi, SB .
LINEAR ALGEBRA AND ITS APPLICATIONS, 2002, 351 :117-145
[8]   CONVERGENCE OF SOME ALGORITHMS FOR CONVEX MINIMIZATION [J].
CORREA, R ;
LEMARECHAL, C .
MATHEMATICAL PROGRAMMING, 1993, 62 (02) :261-275
[9]   IDENTIFYING STRUCTURE OF NONSMOOTH CONVEX FUNCTIONS BY THE BUNDLE TECHNIQUE [J].
Daniilidis, Aris ;
Sagastizabal, Claudia ;
Solodov, Mikhail .
SIAM JOURNAL ON OPTIMIZATION, 2009, 20 (02) :820-840
[10]   Incremental-like bundle methods with application to energy planning [J].
Emiel, Gregory ;
Sagastizabal, Claudia .
COMPUTATIONAL OPTIMIZATION AND APPLICATIONS, 2010, 46 (02) :305-332