The ``fast iterative shrinkage-thresholding algorithm'' (FISTA) is a widely used algorithms in the literature. However, despite its optimal $O(1/k^2)$ theoretical convergence rate, oftentimes its practical performance is not as desired owing to the oscillation behaviour. In this talk, we discuss a simple yet effective modification to the algorithm which has two advantages: 1) it enables us to prove the convergence of the generated sequence; 2) it shows superior practical performance compared to the original FISTA.

This presentation is part of Minisymposium “MS59 - Approaches for fast optimisation in imaging and inverse problems (3 parts)”
organized by: Jingwei Liang (University of Cambridge) , Carola-Bibiane Schönlieb (University of Cambridge) , Mila Nikolova (CMLA - CNRS ENS Cachan, University Paris-Saclay) .

Authors:

Jingwei Liang (University of Cambridge)

Gabriel Peyré (ENS Paris)

Jalal Fadili (Université Caen)

Keywords: