Linear Regression with Unknown Truncation Beyond Gaussian Features

ArXi:2602.12534v2 Announce Type: replace-cross In truncated linear regression, samples $(x,y)$ are shown only when the outcome $y$ falls inside a certain survival set $S^\star$ and the goal is to estimate the unknown $d$-dimensional regressor $w^\star$. This problem has a long history of study in Statistics and Machine Learning going back to the works of (Galton, 1897; Tobin, 1958) and recently in, e.g., (Daskalakis, 2019; 2021; Lee, 2023; 2024). Despite this long history, however, most prior works are limited to the special case where $S^\star$ is precisely known.