{"id":"arxiv/2606.01554","url":"https://arxiv.org/abs/2606.01554","tags":["math.ST"],"score":0,"title":"Fast Near-Optimal Estimation over Symmetric Norm Balls","domain":"math.ST","source":"arxiv","authors":["Matey Neykov"],"reviews":0,"subject":"math.ST","summary":"This short note proposes a polynomial-time algorithm for near-optimal Euclidean estimation of a signal constrained to lie in the unit ball of a symmetric norm, where the symmetry is with respect to a known basis and the norm is accessible through an evaluation oracle. We further extend the method to a random-design, moderate-dimensional linear regression setting, where the regression parameter is likewise assumed to belong to a constraint set defined by a symmetric norm.","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-06-01","reactions":{}}