Averaging on the Bures-Wasserstein manifold: dimension-free convergence of gradient descent
Altschuler, Jason M., Chewi, Sinho, Gerber, Patrik, Stromme, Austin J.
–arXiv.org Artificial Intelligence
We study first-order optimization algorithms for computing the barycenter of Gaussian distributions with respect to the optimal transport metric. Although the objective is geodesically non-convex, Riemannian GD empirically converges rapidly, in fact faster than off-the-shelf methods such as Euclidean GD and SDP solvers. This stands in stark contrast to the best-known theoretical results for Riemannian GD, which depend exponentially on the dimension. In this work, we prove new geodesic convexity results which provide stronger control of the iterates, yielding a dimension-free convergence rate. Our techniques also enable the analysis of two related notions of averaging, the entropically-regularized barycenter and the geometric median, providing the first convergence guarantees for Riemannian GD for these problems.
arXiv.org Artificial Intelligence
Oct-30-2023
- Country:
- North America > United States
- New York (0.04)
- Rhode Island > Providence County
- Providence (0.04)
- New Jersey
- Mercer County > Princeton (0.04)
- Hudson County > Hoboken (0.04)
- Massachusetts
- Middlesex County > Cambridge (0.14)
- Suffolk County > Boston (0.04)
- Europe
- Italy (0.04)
- Switzerland
- Zürich > Zürich (0.04)
- Basel-City > Basel (0.04)
- Asia > Japan
- North America > United States
- Genre:
- Research Report (0.81)
- Industry:
- Government > Regional Government (0.45)
- Technology: