Rejoinder to Letter to the Editors “Comments on ‘Statistical inference on change points in generalized semiparametric segmented models’ by Yang et al. (2025)” by Vito M.R. Muggeo

Résumé fourni par la source

We would like to thank Dr Muggeo for his thoughtful letter and influential contributions to the field, particularly through the development of the widely used R package “segmented”. Change-point estimation has wide-ranging applications in medicine, biology, environmental studies, and many other domains. At the same time, the methodology continues to pose challenging theoretical questions and practical issues, underscoring the need for further rigorous research. We welcome this exchange and now address the specific points raised in the letter. 1. On the convergence properties of the ToSNR algorithm In Paragraph 3 of his letter, Dr. Muggeo correctly noted that the objective function in segmented regression models can exhibit multiple stationary points, and thus iterative ToSNR algorithms may converge to local rather than global optima. He further suggested practical remedies, such as adopting multiple starting values or bootstrap resampling, as implemented in the widely used R package “segmented” (Muggeo et al., 2008).

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.