Dual Prices for Frank--Wolfe Algorithms
Optimization and Control
2021-02-08 v2
Abstract
In this note we observe that for constrained convex minimization problems over a polytope , dual prices for the linear program obtained from linearization at approximately optimal solutions have a similar interpretation of rate of change in optimal value as for linear programming, providing a convex form of sensitivity analysis. This is of particular interest for Frank--Wolfe algorithms (also called conditional gradients), forming an important class of first-order methods, where a basic building block is linear minimization of gradients of over , which in most implementations already compute the dual prices as a by-product.
Cite
@article{arxiv.2101.02087,
title = {Dual Prices for Frank--Wolfe Algorithms},
author = {Gábor Braun and Sebastian Pokutta},
journal= {arXiv preprint arXiv:2101.02087},
year = {2021}
}
Comments
4 pages; short note (added funding acknowledgement in v2)