Minimum Rényi pseudodistance estimators for logistic regression models
No Thumbnail Available
Official URL
Full text at PDC
Publication date
2022
Advisors (or tutors)
Journal Title
Journal ISSN
Volume Title
Publisher
Springer
Citation
Alonso, J. M., Calviño, A., & Muñoz, S. (2022). Minimum Rényi Pseudodistance Estimators for Logistic Regression Models. Trends in Mathematical, Information and Data Sciences: A Tribute to Leandro Pardo, 131-145.
Abstract
In this work we propose a new family of estimators, called minimum Rényi pseudodistance estimators (MRPE), as a robust generalization of maximum likelihood estimators (MLE) for the logistic regression model based on the Rényi pseudodistance introduced by Jones et al. [14], along with their corresponding asymptotic distribution. Based on this information, we further develop three types of confidence intervals (approximate and parametric and non-parametric bootstrap ones). Finally, a simulation study is conducted considering different levels of outliers, where a better behavior of the MRPE with respect to the MLE is shown.