Convergence and complexity of block majorization-minimization for constrained block-Riemannian optimization
Abstract
Block majorization-minimization (BMM) is a simple iterative algorithm for nonconvex optimization that sequentially minimizes a majorizing surrogate of the objective function in each block coordinate while the other block coordinates are held fixed. We consider a family of BMM algorithms for minimizing smooth nonconvex objectives, where each parameter block is constrained within a subset of a Riemannian manifold. We establish that this algorithm converges asymptotically to the set of stationary points, and attains an -stationary point within iterations. In particular, the assumptions for our complexity results are completely Euclidean when the underlying manifold is a product of Euclidean or Stiefel manifolds, although our analysis makes explicit use of the Riemannian geometry. Our general analysis applies to a wide range of algorithms with Riemannian constraints: Riemannian MM, block projected gradient descent, optimistic likelihood estimation, geodesically constrained subspace tracking, robust PCA, and Riemannian CP-dictionary-learning. We experimentally validate that our algorithm converges faster than standard Euclidean algorithms applied to the Riemannian setting.
Cite
@article{arxiv.2312.10330,
title = {Convergence and complexity of block majorization-minimization for constrained block-Riemannian optimization},
author = {Yuchen Li and Laura Balzano and Deanna Needell and Hanbaek Lyu},
journal= {arXiv preprint arXiv:2312.10330},
year = {2026}
}
Comments
54 pages, 8 figures. Related work updated