{"id":"arxiv/2604.14017","url":"https://arxiv.org/abs/2604.14017","tags":["math.OC","cs.LG"],"score":0,"title":"Stochastic Trust-Region Methods for Over-parameterized Models","domain":"math.OC","source":"arxiv","authors":["Aike Yang","Hao Wang"],"reviews":0,"subject":"math.OC","summary":"Under interpolation-type assumptions such as the strong growth condition, stochastic optimization methods can attain convergence rates comparable to full-batch methods, but their performance, particularly for SGD, remains highly sensitive to step-size selection. To address this issue, we propose a unified stochastic trust-region framework that eliminates manual step-size tuning and extends naturally to equality-constrained problems. For unconstrained optimization, we develop a first-order stochastic trust-region algorithm and show that, under the strong growth condition, it achieves an…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-04-15","reactions":{}}