INCREMENTAL SUBGRADIENTS FOR CONSTRAINED CONVEX OPTIMIZATION: A UNIFIED FRAMEWORK AND NEW METHODS

Autor(es)

Helou Neto, Elias Salomao ^[1] ; De Pierro, Alvaro Rodolfo ^[1]

Afiliação do(s) autor(es)

^[1] Univ Estadual Campinas, Dept Appl Math, BR-13081970 Campinas, SP - Brazil

• Tipo de documento: Artigo Científico
• Fonte: SIAM JOURNAL ON OPTIMIZATION; v. 20, n. 3, p. 1547-1572, 2009.
• Citações Web of Science: 23

Resumo

We present a unifying framework for nonsmooth convex minimization bringing together is an element of-subgradient algorithms and methods for the convex feasibility problem. This development is a natural step for is an element of-subgradient methods in the direction of constrained optimization since the Euclidean projection frequently required in such methods is replaced by an approximate projection, which is often easier to compute. The developments are applied to incremental subgradient methods, resulting in new algorithms suitable to large-scale optimization problems, such as those arising in tomographic imaging. (AU)

• Processo FAPESP: 02/07153-2 - Algoritmos para a reconstrução tomográfica: otimização, restauração, quantificação e aplicação
• Beneficiário: Sergio Shiguemi Furuie
• Modalidade de apoio: Auxílio à Pesquisa - Temático