Bregman Proximal Point Algorithm Revisited: A New Inexact Version and its Inertial Variant

TL;DR

This paper revisits the Bregman proximal point algorithm, introduces a flexible inexact stopping condition, and develops an inertial variant with improved convergence rates for convex optimization problems, including optimal transport.

Contribution

It proposes a new inexact stopping condition for BPPA, making it more adaptable, and introduces an inertial variant with accelerated convergence, supported by theoretical analysis and preliminary experiments.

Findings

The inexact BPPA covers existing conditions as special cases.

The inertial V-iBPPA achieves an $O(1/k^2)$ convergence rate under certain conditions.

Preliminary experiments demonstrate improved convergence speed of V-iBPPA.

Abstract

We study a general convex optimization problem, which covers various classic problems in different areas and particularly includes many optimal transport related problems arising in recent years. To solve this problem, we revisit the classic Bregman proximal point algorithm (BPPA) and introduce a new inexact stopping condition for solving the subproblems, which can circumvent the underlying feasibility difficulty often appearing in existing inexact conditions when the problem has a complex feasible set. Our inexact condition also covers several existing inexact conditions as special cases and hence makes our inexact BPPA (iBPPA) more flexible to fit different scenarios in practice. Moreover, inspired by Nesterov's acceleration technique, we develop an inertial variant of our iBPPA, denoted by V-iBPPA, and establish the iteration complexity of $O(1/k^{\lambda})$, where $\lambda\geq1$ is a quadrangle scaling exponent of the kernel function. In particular, when the proximal parameter is a constant and the kernel function is strongly convex with Lipschitz continuous gradient (hence $\lambda=2$), our V-iBPPA achieves a faster rate of $O(1/k^2)$ just as existing accelerated inexact proximal point algorithms. Some preliminary numerical experiments for solving the standard OT problem are conducted to show the convergence behaviors of our iBPPA and V-iBPPA under different inexactness settings. The experiments also empirically verify the potential of our V-iBPPA on improving the convergence speed.