caa a22 4500 38807101X CHVBK 20191024065209.0 cr unu---uuuuu 161130e199911 xx s 000 0 eng 10.2143/AST.29.2.504614 doi S0515036100012575 pii (NATIONALLICENCE)cambridge-10.2143/AST.29.2.504614 Selection of Credibility Regression Models [Elektronische Daten] We derive some decision rules to select best predictive regression models in a credibility context, that is, in a "random effects' linear regression model with replicates. In contrast to usual model selection techniques on a collective level, our proposal allows to detect individual structures, even if they disappear in the collective. We give exact, non-asymptotic results for the expected squared error loss for a predictor based on credibility estimation in different models. This involves correct accounting of random model parameters and the study of expected loss for shrinkage estimation. We support the theoretical properties of the new model selectors by a small simulation experiment. Copyright © International Actuarial Association 1999 Bias-variance trade-off nationallicence Empirical Bayes nationallicence Random effects model nationallicence Shrinkage estimation nationallicence Squared prediction loss nationallicence Subset-regression nationallicence Bühlmann Peter ETH Zürich, Switzerland Bühlmann Hans ETH Zürich, Switzerland ASTIN Bulletin Cambridge University Press 29/2(1999-11), 245-270 0515-0361 29:2<245 1999 29 ASB https://doi.org/10.2143/AST.29.2.504614 text/html Onlinezugriff via DOI 1 research-article jats NATIONALLICENCE 856 40 https://doi.org/10.2143/AST.29.2.504614 text/html Onlinezugriff via DOI NATIONALLICENCE 700 1- Bühlmann Peter ETH Zürich, Switzerland NATIONALLICENCE 700 1- Bühlmann Hans ETH Zürich, Switzerland NATIONALLICENCE 773 0- ASTIN Bulletin Cambridge University Press 29/2(1999-11), 245-270 0515-0361 29:2<245 1999 29 ASB CC0 http://creativecommons.org/publicdomain/zero/1.0 nationallicence BK010053 XK010053 XK010000 NATIONALLICENCE NATIONALLICENCE NL-cambridge