Iteration-Complexity of a Generalized Forward Backward Splitting Algorithm

In this paper, we analyze the iteration-complexity of a generalized forward-backward (GFB) splitting algorithm, recently proposed in~\cite{gfb2011}, for minimizing the large class of composite objectives $f + \sum_{i=1}^n h_i$ on a Hilbert space, where $f$ has a Lipschitz-continuous gradient and the $h_i$'s are simple (i.e. whose proximity operator is easily computable ). We derive iteration-complexity bounds (pointwise and ergodic) for GFB to obtain an approximate solution based on an easily verifiable termination criterion. Along the way, we prove complexity bounds for relaxed fixed point iterations built from composition of nonexpansive averaged operators. These results apply more generally to GFB when used to find a zero of a sum of $n > 0$ maximal monotone operators and a co-coercive operator on a Hilbert space. The theoretical findings are exemplified with experiments on video processing.

Citation

GREYC CNRS-ENSICAEN-University of Caen, 6, Bd du Maréchal Juin, 14050 Caen Cedex, France, 10/2013

Article

Download

View Iteration-Complexity of a Generalized Forward Backward Splitting Algorithm