Honest Confidence Regions for Logistic Regression with a Large Number of ControlsReport as inadecuate



 Honest Confidence Regions for Logistic Regression with a Large Number of Controls


Honest Confidence Regions for Logistic Regression with a Large Number of Controls - Download this document for free, or read online. Document in PDF available to download.

Download or read this book online for free in PDF: Honest Confidence Regions for Logistic Regression with a Large Number of Controls
This paper considers inference in logistic regression models with high dimensional data. We propose new inference tools on constructing confidence regions for a regression parameter of primary interest $\alpha 0$. These tools allow the total number of controls to exceed the sample size when only a subset of the controls are needed for the regression. Importantly, we show that these resulting confidence regions are -honest- in the formal sense that they hold uniformly over many data-generating processes, and do not rely on traditional consistent model selection arguments. Consequently, the inferential results derived are robust against model selection mistakes.



Author: Alexandre Belloni; Victor Chernozhukov; Ying Wei

Source: https://archive.org/







Related documents