Global rates of convergence for nonconvex optimization on manifolds
Abstract
We consider the minimization of a cost function on a manifold using Riemannian gradient descent and Riemannian trust regions (RTR). We focus on satisfying necessary optimality conditions within a tolerance . Specifically, we show that, under Lipschitz-type assumptions on the pullbacks of to the tangent spaces of , both of these algorithms produce points with Riemannian gradient smaller than in iterations. Furthermore, RTR returns a point where also the Riemannian Hessian's least eigenvalue is larger than in iterations. There are no assumptions on initialization. The rates match their (sharp) unconstrained counterparts as a function of the accuracy (up to constants) and hence are sharp in that sense. These are the first deterministic results for global rates of convergence to approximate first- and second-order Karush-Kuhn-Tucker points on manifolds. They apply in particular for optimization constrained to compact submanifolds of , under simpler assumptions.
Keywords
Cite
@article{arxiv.1605.08101,
title = {Global rates of convergence for nonconvex optimization on manifolds},
author = {Nicolas Boumal and P. -A. Absil and Coralia Cartis},
journal= {arXiv preprint arXiv:1605.08101},
year = {2018}
}
Comments
33 pages, IMA Journal of Numerical Analysis, 2018