Convergence rate analysis of a Dykstra-type projection algorithm

TL;DR

This paper extends Dykstra's projection algorithm to solve the best approximation problem in multi-set split feasibility settings involving convex sets and linear maps, providing explicit convergence rates under certain geometric conditions.

Contribution

It introduces a Dykstra-type algorithm for the general BA-MSF problem using dual coordinate gradient descent and establishes explicit convergence rates based on the Kurdyka-Lojasiewicz property.

Findings

The algorithm converges with explicit rates under standard conditions.

Convergence can be linear or sublinear depending on set regularity.

Examples demonstrate the necessity of assumptions.

Abstract

Given closed convex sets $C_i$, $i=1,\ldots,\ell$, and some nonzero linear maps $A_i$, $i = 1,\ldots,\ell$, of suitable dimensions, the multi-set split feasibility problem aims at finding a point in $\bigcap_{i=1}^\ell A_i^{-1}C_i$ based on computing projections onto $C_i$ and multiplications by $A_i$ and $A_i^T$. In this paper, we consider the associated best approximation problem, i.e., the problem of computing projections onto $\bigcap_{i=1}^\ell A_i^{-1}C_i$; we refer to this problem as the best approximation problem in multi-set split feasibility settings (BA-MSF). We adapt the Dykstra's projection algorithm, which is classical for solving the BA-MSF in the special case when all $A_i = I$, to solve the general BA-MSF. Our Dykstra-type projection algorithm is derived by applying (proximal) coordinate gradient descent to the Lagrange dual problem, and it only requires computing projections onto $C_i$ and multiplications by $A_i$ and $A_i^T$ in each iteration. Under a standard relative interior condition and a genericity assumption on the point we need to project, we show that the dual objective satisfies the Kurdyka-Lojasiewicz property with an explicitly computable exponent on a neighborhood of the (typically unbounded) dual solution set when each $C_i$ is $C^{1,\alpha}$-cone reducible for some $\alpha\in (0,1]$: this class of sets covers the class of $C^2$-cone reducible sets, which include all polyhedrons, second-order cone, and the cone of positive semidefinite matrices as special cases. Using this, explicit convergence rate (linear or sublinear) of the sequence generated by the Dykstra-type projection algorithm is derived. Concrete examples are constructed to illustrate the necessity of some of our assumptions.