@article{Klein2015BayesianStructured,
 abstract = {We propose a generic Bayesian framework for inference in distributional

regression models in which each parameter of a potentially complex response

distribution and not only the mean is related to a structured additive

predictor. The latter is composed additively of a variety of different

functional effect types such as nonlinear effects, spatial effects, random

coefficients, interaction surfaces or other (possibly nonstandard) basis

function representations. To enforce specific properties of the functional

effects such as smoothness, informative multivariate Gaussian priors are

assigned to the basis function coefficients. Inference can then be based on

computationally efficient Markov chain Monte Carlo simulation techniques where

a generic procedure makes use of distribution-specific iteratively weighted

least squares approximations to the full conditionals. The framework of

distributional regression encompasses many special cases relevant for treating

nonstandard response structures such as highly skewed nonnegative responses,

overdispersed and zero-inflated counts or shares including the possibility for

zero- and one-inflation. We discuss distributional regression along a study on

determinants of labour incomes for full-time working males in Germany with a

particular focus on regional differences after the German reunification.

Controlling for age, education, work experience and local disparities, we

estimate full conditional income distributions allowing us to study various

distributional quantities such as moments, quantiles or inequality measures in

a consistent manner in one joint model. Detailed guidance on practical aspects

of model choice including the selection of several competing distributions for

labour incomes and the consideration of different covariate effects on the

income distribution complete the distributional regression analysis. We find

that next to a lower expected income, full-time working men in East Germany

also face a more unequal income distribution than men in the West, ceteris

paribus.},
 author = {Klein, Nadja and Kneib, Thomas and Lang, Stefan and Sohn, Alexander},
 year = {2015},
 title = {Bayesian structured additive distributional regression with an application to regional income inequality in {Germany}},
 url = {http://dx.doi.org/10.1214/15-aoas823},
 keywords = {phd;stat},
 pages = {1024--1052},
 volume = {9},
 number = {2},
 issn = {1932-6157},
 journal = {The Annals of Applied Statistics},
 doi = {10.1214/15-aoas823},
 howpublished = {refereed}
}

{"_id":"wc2hFGXv8GbNXKBvB","bibbaseid":"klein-kneib-lang-sohn-bayesianstructuredadditivedistributionalregressionwithanapplicationtoregionalincomeinequalityingermany-2015","downloads":0,"creationDate":"2019-02-27T08:30:59.449Z","title":"Bayesian structured additive distributional regression with an application to regional income inequality in Germany","author_short":["Klein, N.","Kneib, T.","Lang, S.","Sohn, A."],"year":2015,"bibtype":"article","biburl":"http://www.uni-goettingen.de/de/document/download/9d7c40531010bf5be953ccd9446e47ae.bib/GRK1644BibHomepage.bib","bibdata":{"bibtype":"article","type":"article","abstract":"We propose a generic Bayesian framework for inference in distributional regression models in which each parameter of a potentially complex response distribution and not only the mean is related to a structured additive predictor. The latter is composed additively of a variety of different functional effect types such as nonlinear effects, spatial effects, random coefficients, interaction surfaces or other (possibly nonstandard) basis function representations. To enforce specific properties of the functional effects such as smoothness, informative multivariate Gaussian priors are assigned to the basis function coefficients. Inference can then be based on computationally efficient Markov chain Monte Carlo simulation techniques where a generic procedure makes use of distribution-specific iteratively weighted least squares approximations to the full conditionals. The framework of distributional regression encompasses many special cases relevant for treating nonstandard response structures such as highly skewed nonnegative responses, overdispersed and zero-inflated counts or shares including the possibility for zero- and one-inflation. We discuss distributional regression along a study on determinants of labour incomes for full-time working males in Germany with a particular focus on regional differences after the German reunification. Controlling for age, education, work experience and local disparities, we estimate full conditional income distributions allowing us to study various distributional quantities such as moments, quantiles or inequality measures in a consistent manner in one joint model. Detailed guidance on practical aspects of model choice including the selection of several competing distributions for labour incomes and the consideration of different covariate effects on the income distribution complete the distributional regression analysis. We find that next to a lower expected income, full-time working men in East Germany also face a more unequal income distribution than men in the West, ceteris paribus.","author":[{"propositions":[],"lastnames":["Klein"],"firstnames":["Nadja"],"suffixes":[]},{"propositions":[],"lastnames":["Kneib"],"firstnames":["Thomas"],"suffixes":[]},{"propositions":[],"lastnames":["Lang"],"firstnames":["Stefan"],"suffixes":[]},{"propositions":[],"lastnames":["Sohn"],"firstnames":["Alexander"],"suffixes":[]}],"year":"2015","title":"Bayesian structured additive distributional regression with an application to regional income inequality in Germany","url":"http://dx.doi.org/10.1214/15-aoas823","keywords":"phd;stat","pages":"1024–1052","volume":"9","number":"2","issn":"1932-6157","journal":"The Annals of Applied Statistics","doi":"10.1214/15-aoas823","howpublished":"refereed","bibtex":"@article{Klein2015BayesianStructured,\r\n abstract = {We propose a generic Bayesian framework for inference in distributional\r\n\r\nregression models in which each parameter of a potentially complex response\r\n\r\ndistribution and not only the mean is related to a structured additive\r\n\r\npredictor. The latter is composed additively of a variety of different\r\n\r\nfunctional effect types such as nonlinear effects, spatial effects, random\r\n\r\ncoefficients, interaction surfaces or other (possibly nonstandard) basis\r\n\r\nfunction representations. To enforce specific properties of the functional\r\n\r\neffects such as smoothness, informative multivariate Gaussian priors are\r\n\r\nassigned to the basis function coefficients. Inference can then be based on\r\n\r\ncomputationally efficient Markov chain Monte Carlo simulation techniques where\r\n\r\na generic procedure makes use of distribution-specific iteratively weighted\r\n\r\nleast squares approximations to the full conditionals. The framework of\r\n\r\ndistributional regression encompasses many special cases relevant for treating\r\n\r\nnonstandard response structures such as highly skewed nonnegative responses,\r\n\r\noverdispersed and zero-inflated counts or shares including the possibility for\r\n\r\nzero- and one-inflation. We discuss distributional regression along a study on\r\n\r\ndeterminants of labour incomes for full-time working males in Germany with a\r\n\r\nparticular focus on regional differences after the German reunification.\r\n\r\nControlling for age, education, work experience and local disparities, we\r\n\r\nestimate full conditional income distributions allowing us to study various\r\n\r\ndistributional quantities such as moments, quantiles or inequality measures in\r\n\r\na consistent manner in one joint model. Detailed guidance on practical aspects\r\n\r\nof model choice including the selection of several competing distributions for\r\n\r\nlabour incomes and the consideration of different covariate effects on the\r\n\r\nincome distribution complete the distributional regression analysis. We find\r\n\r\nthat next to a lower expected income, full-time working men in East Germany\r\n\r\nalso face a more unequal income distribution than men in the West, ceteris\r\n\r\nparibus.},\r\n author = {Klein, Nadja and Kneib, Thomas and Lang, Stefan and Sohn, Alexander},\r\n year = {2015},\r\n title = {Bayesian structured additive distributional regression with an application to regional income inequality in {Germany}},\r\n url = {http://dx.doi.org/10.1214/15-aoas823},\r\n keywords = {phd;stat},\r\n pages = {1024--1052},\r\n volume = {9},\r\n number = {2},\r\n issn = {1932-6157},\r\n journal = {The Annals of Applied Statistics},\r\n doi = {10.1214/15-aoas823},\r\n howpublished = {refereed}\r\n}\r\n\r\n\r\n","author_short":["Klein, N.","Kneib, T.","Lang, S.","Sohn, A."],"key":"Klein2015BayesianStructured","id":"Klein2015BayesianStructured","bibbaseid":"klein-kneib-lang-sohn-bayesianstructuredadditivedistributionalregressionwithanapplicationtoregionalincomeinequalityingermany-2015","role":"author","urls":{"Paper":"http://dx.doi.org/10.1214/15-aoas823"},"keyword":["phd;stat"],"metadata":{"authorlinks":{}},"downloads":0},"search_terms":["bayesian","structured","additive","distributional","regression","application","regional","income","inequality","germany","klein","kneib","lang","sohn"],"keywords":["phd;stat"],"authorIDs":[],"dataSources":["psxr4mFyE5JDwFLuZ","2w3D54bmLuhpt4TNv","t8S6Y6RWEwDAQHiSQ","cLGdYAfLyvQDgrYmh"]}