Single Parameter Inference of Non-sparse Logistic Regression Models

TL;DR

This paper introduces a method for inferring a single parameter in non-sparse logistic regression models by transforming the null hypothesis into a moment condition, with demonstrated good performance through numerical experiments.

Contribution

It presents a novel approach that transforms the null hypothesis into a moment condition for inference in non-sparse logistic regression models.

Findings

Method performs well in numerical experiments

Constructs test statistic with asymptotic null distribution

Effective for single parameter inference in non-sparse models

Abstract

This paper infers a single parameter in non-sparse logistic regression models. By transforming the null hypothesis into a moment condition, we construct the test statistic and obtain the asymptotic null distribution. Numerical experiments show that our method performs well.