{"id":"arxiv/2610.19944","url":"https://arxiv.org/abs/2610.19944","tags":["math.OC"],"score":0,"title":"Recursive Gradient Retraction Methods for Stochastic Nonconvex Optimization on Stiefel Manifolds","domain":"math.OC","source":"arxiv","authors":["Zhijie Yuan","Lei Sun","Ganzhao Yuan"],"reviews":0,"subject":"math.OC","summary":"Nonconvex optimization on Stiefel manifolds underlies many problems in machine learning and deep learning. However, existing stochastic Riemannian retraction methods often rely on stepsizes that require prior knowledge of the gradient Lipschitz constant, a prescribed iteration horizon, or restrictive batch-size choices. We develop {\\sf RGR}, a recursive gradient retraction method for stochastic nonconvex optimization on Stiefel manifolds. Instead of recursively updating gradient estimators in changing tangent spaces, {\\sf RGR} constructs them in the Euclidean space and subsequently maps them…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-10-14","reactions":{}}