Nonparametric Instrumental Variable Regression with Observed Covariates
Zikai Shen ∗, Zonghao Chen ∗, Dimitri Meunier, Ingo Steinwart, Arthur Gretton †, Zhu Li † · 2025
Read Nonparametric Instrumental Variable Regression with Observed Covariates by Zikai Shen ∗, Zonghao Chen ∗, Dimitri Meunier, Ingo Steinwart, Arthur Gretton †, Zhu Li † (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-19404.txt.
Introduction
Zikai Shen and five co-authors extend instrumental variable regression to settings where extra covariates are observed directly. Instruments help untangle cause and effect when a key driver is hidden, and observed covariates can sharpen identification and reveal varied effects. They also break the assumptions of older theory in two ways.
The covariates create a partial identity structure and uneven smoothness across directions. The authors add a Fourier measure of partial smoothing and adapt a kernel two-stage method with direction-aware length scales. They prove learning rate bounds from above and the first minimax lower bounds for this problem, which bridge the known optimal rates.
Catalog record
| CREATOR | Zikai Shen ∗, Zonghao Chen ∗, Dimitri Meunier, Ingo Steinwart, Arthur Gretton †, Zhu Li † |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 337 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-19404