{"id":"arxiv/2610.14414","url":"https://arxiv.org/abs/2610.14414","tags":["stat.ME","math.ST","stat.ML"],"score":0,"title":"Prior-Informed Generalizable Low-Rank Representation Learning via Regularized Adversarial Reweighting","domain":"stat.ME","source":"arxiv","authors":["Keyao Zhan","Junwei Lu","Tianxi Cai"],"reviews":0,"subject":"stat.ME","summary":"Modern learning problems increasingly leverage data from heterogeneous sources to improve robustness and generalizability under distribution shift, often with high-dimensional parameters admitting low-rank structure. Existing group distributionally robust optimization (group-DRO) approaches optimize against adverse source mixtures, but extending this principle to low-rank matrix estimation with a broad class of losses leads to a challenging nonconvex-concave problem for which methods and theory remain limited. We propose Prior-Informed generalizable Low-rank representation Learning via…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-10-10","reactions":{}}