Scholarlink

<p>Change-point analysis for logistic regression with binary outcomes is pivotal in areas where a latent threshold can alter risk relationships. We introduce the heterogeneous change-point logistic regression, allowing each individual’s change-point to be a random variable rather than a fixed threshold. The model estimation is carried out via an expectation conditional maximization algorithm. The theoretical properties of the proposed estimators are also studied. To detect the existence of the structural change-point and its randomness, we propose a supremum-type (SUP) test and a likelihood ratio test, respectively. Simulation results illustrate the good performance of our methods in the finite sample. We also analyze a school mental-health screening dataset collected from junior high school students in China, which shows that the change-points emerge in academic anxiety.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Heterogeneous change-point logistic regression for binary school mental-health alerts

  • 期刊论文

  • 发表日期：

    2026年5月19日

查看全文
  • Ji’an Lei,
  • Qing Jiang,
  • Yunhang Guo,
  • Xingwei Tong,
  • Nan Zhao

摘要

Change-point analysis for logistic regression with binary outcomes is pivotal in areas where a latent threshold can alter risk relationships. We introduce the heterogeneous change-point logistic regression, allowing each individual’s change-point to be a random variable rather than a fixed threshold. The model estimation is carried out via an expectation conditional maximization algorithm. The theoretical properties of the proposed estimators are also studied. To detect the existence of the structural change-point and its randomness, we propose a supremum-type (SUP) test and a likelihood ratio test, respectively. Simulation results illustrate the good performance of our methods in the finite sample. We also analyze a school mental-health screening dataset collected from junior high school students in China, which shows that the change-points emerge in academic anxiety.