Minimum Norm Estimation Under Parameter Constraints with an Application to Insurance

Minimum Norm Estimation Under Parameter Constraints with an Application to Insurance

A Minimum Norm Quadratic Estimator is developed for situations where some fixed effects and variance components coincide. The study is motivated by a semiparametric latent variable model frequently encountered in insurance, where a Poisson assumption induces identity between the mean and the within-...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Statistics (Berlin, DDR) DDR), 1998-01, Vol.31 (3), p.215-234
Hauptverfasser: Kleffe, Jürgen, Norberg, Ragnar
Format: Artikel
Sprache:eng
Schlagworte:
62C12
empirical Bayes
linear models
Variance components
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:A Minimum Norm Quadratic Estimator is developed for situations where some fixed effects and variance components coincide. The study is motivated by a semiparametric latent variable model frequently encountered in insurance, where a Poisson assumption induces identity between the mean and the within-unit variance. The method is applied to authentic group life insurance data, and its performance is also illustrated by simulations.
ISSN:0233-1888
1029-4910
DOI:10.1080/02331889808802637