期刊
SIAM JOURNAL ON OPTIMIZATION
卷 29, 期 1, 页码 541-570出版社
SIAM PUBLICATIONS
DOI: 10.1137/17M1124085
关键词
abstract convergence theorem; Kurdyka-Lojasiewicz inequality; descent method; relative errors; block coordinate method; variable metric method; inertial method; iPiano; inpainting; Mumford-Shah regularizer
An abstract convergence theorem for a class of generalized descent methods that explicitly models relative errors is proved. The convergence theorem generalizes and unifies several recent abstract convergence theorems. It is applicable to possibly nonsmooth and nonconvex lower semicontinuous functions that satisfy the Kurdyka-Lojasiewicz (KL) inequality, which comprises a huge class of problems. Many of the recent algorithms that explicitly prove convergence using the KL inequality can be cast in the abstract framework of this paper and, therefore, the generated sequence converges to a stationary point of the objective function. Additional flexibility compared to related approaches is gained by a descent property that is formulated with respect to a function that is allowed to change along the iterations, a generic distance measure, and an explicit/implicit relative error condition with respect to finite linear combinations of distance terms. As an application of the gained flexibility, the convergence of a block coordinate variable metric version of iPiano (an inertial forward-backward splitting algorithm) is proved, which performs favorably on an inpainting problem with a Mumford-Shah-like regularization from image processing.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据