作者: Patrick L Combettes , Bằng C Vũ , None
关键词: Convex optimization 、 Proximal Gradient Methods 、 Convex metric space 、 Norm (mathematics) 、 Monotonic function 、 Injective metric space 、 Intrinsic metric 、 Mathematics 、 Mathematical optimization 、 Landweber iteration
摘要: The notion of quasi-Fejer monotonicity has proven to be an efficient tool simplify and unify the convergence analysis various algorithms arising in applied nonlinear analysis. In this paper, we extend context variable metric algorithms, whereby underlying norm is allowed vary at each iteration. Applications convex optimization inverse problems are demonstrated.