Averaging on the Bures-Wasserstein manifold: dimension-free convergence of gradient descent
–Neural Information Processing Systems
We study first-order optimization algorithms for computing the barycenter of Gaussian distributions with respect to the optimal transport metric.
Neural Information Processing Systems
Aug-17-2025, 02:22:51 GMT
- Country:
- Asia > Russia (0.04)
- Europe
- Italy (0.04)
- Russia (0.04)
- Switzerland
- Basel-City > Basel (0.04)
- Zürich > Zürich (0.04)
- North America > United States
- Massachusetts
- Middlesex County > Cambridge (0.04)
- Suffolk County > Boston (0.04)
- New Jersey
- Hudson County > Hoboken (0.04)
- Mercer County > Princeton (0.04)
- Rhode Island > Providence County
- Providence (0.04)
- Massachusetts
- Industry:
- Government > Regional Government (0.45)
- Technology: