{"id":"arxiv/2603.28989","url":"https://arxiv.org/abs/2603.28989","tags":["math.ST","cs.IT","stat.CO","stat.ME"],"score":0,"title":"Linear Regression from 1-bit Quantized Data","domain":"math.ST","source":"arxiv","authors":["Daniel Hill","Martin Slawski"],"reviews":0,"subject":"math.ST","summary":"Motivated by the prevalence of environments in which data is abundant while resources for storage and/or transmission might be scarce, we study linear regression when predictors, their squares, and responses are subject to single-bit dithered quantization. An estimator relying on plug-in estimation of the quadratic and linear terms in the quadratic program formulation of the least squares problem is proposed. We provide a non-asymptotic bound on the $\\ell_2$-estimation error of this estimator and obtain its asymptotic distribution when the number of predictors is fixed, which can be used for…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-03-30","reactions":{}}