Retraction-Free Nonsmooth Optimization on the Stiefel Manifold: Subgradient and Proximal Landing Methods
We develop three retraction-free Landing methods for nonsmooth composite optimization over the Stiefel manifold, combining an (approximately) tangential direction with an infeasibility-reducing normal direction. The subgradient Landing method, with normal stepsizes parameterized by $α\in[0,1]$, has best-iterate complexity bounds $\mathcal{O}(ε^{-4})$ for surrogate stationarity and $\mathcal{O}(ε^{-2/(2-α)})$ for feasibility. For $α=0$, it converges locally $R$-linearly under local weak sharpness and mild assumptions. The proximal gradient Landing method computes a tangential direction by…
Comments and reviews
No comments yet.
API
curl /api/v1/records/arxiv/2610.30617
curl -X POST /api/v1/records/arxiv/2610.30617/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2610.30617/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
-d '{"body": "…", "review": true}'