Outlier identification rules for generalized linear models

S. Kuhnt; J. Pawlitschko

doi:10.1007/3-540-26981-9_20

Outlier identification rules for generalized linear models

S. Kuhnt, J. Pawlitschko

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/Congresprocedure › Conferentiebijdrage › Academic › peer review

Samenvatting

Observations which seem to deviate strongly from the main part of the data may occur in every statistical analysis. These observations, usually labelled as outliers, may cause completely misleading results when using standard methods and may also contain information about special events or dependencies. We discuss outliers in situations where a generalized linear model is assumed as null model for the regular data and introduce rules for their identification. For the special cases of a loglinear Poisson model and a logistic regression model some one-step identifiers based on robust and non-robust estimators are proposed and compared.

Originele taal-2	Engels
Titel	Innovations in Classification, Data Science, and Information Systems (Proceedings of the 27th Annual Conference of the Gesellschaft für Klassifikation e.V., Cottbus, Germany, March 12–14, 2003), Part II
Redacteuren	D. Baier, K.D. Warnecke
Plaats van productie	Berlin
Uitgeverij	Springer
Hoofdstuk	20
Pagina's	165-172
Aantal pagina's	8
ISBN van elektronische versie	978-3-540-26981-6
ISBN van geprinte versie	3-540-23221-4, 978-3-540-23221-6
DOI's	https://doi.org/10.1007/3-540-26981-9_20
Status	Gepubliceerd - 2005

Publicatie series

Naam	Studies in Classification, Data Analysis, and Knowledge Organization

Toegang tot document

10.1007/3-540-26981-9_20

Citeer dit

Kuhnt, S., & Pawlitschko, J. (2005). Outlier identification rules for generalized linear models. In D. Baier, & K. D. Warnecke (editors), Innovations in Classification, Data Science, and Information Systems (Proceedings of the 27th Annual Conference of the Gesellschaft für Klassifikation e.V., Cottbus, Germany, March 12–14, 2003), Part II (blz. 165-172). (Studies in Classification, Data Analysis, and Knowledge Organization). Springer. https://doi.org/10.1007/3-540-26981-9_20

Kuhnt, S. ; Pawlitschko, J. / Outlier identification rules for generalized linear models. Innovations in Classification, Data Science, and Information Systems (Proceedings of the 27th Annual Conference of the Gesellschaft für Klassifikation e.V., Cottbus, Germany, March 12–14, 2003), Part II. redacteur / D. Baier ; K.D. Warnecke. Berlin : Springer, 2005. blz. 165-172 (Studies in Classification, Data Analysis, and Knowledge Organization).

@inproceedings{c46bdeeb436341b6989dfec9a6a27cb3,

title = "Outlier identification rules for generalized linear models",

abstract = "Observations which seem to deviate strongly from the main part of the data may occur in every statistical analysis. These observations, usually labelled as outliers, may cause completely misleading results when using standard methods and may also contain information about special events or dependencies. We discuss outliers in situations where a generalized linear model is assumed as null model for the regular data and introduce rules for their identification. For the special cases of a loglinear Poisson model and a logistic regression model some one-step identifiers based on robust and non-robust estimators are proposed and compared.",

author = "S. Kuhnt and J. Pawlitschko",

year = "2005",

doi = "10.1007/3-540-26981-9_20",

language = "English",

isbn = "3-540-23221-4",

series = "Studies in Classification, Data Analysis, and Knowledge Organization",

publisher = "Springer",

pages = "165--172",

editor = "D. Baier and K.D. Warnecke",

booktitle = "Innovations in Classification, Data Science, and Information Systems (Proceedings of the 27th Annual Conference of the Gesellschaft f{\"u}r Klassifikation e.V., Cottbus, Germany, March 12–14, 2003), Part II",

address = "Germany",

}

Kuhnt, S & Pawlitschko, J 2005, Outlier identification rules for generalized linear models. in D Baier & KD Warnecke (uitgave), Innovations in Classification, Data Science, and Information Systems (Proceedings of the 27th Annual Conference of the Gesellschaft für Klassifikation e.V., Cottbus, Germany, March 12–14, 2003), Part II. Studies in Classification, Data Analysis, and Knowledge Organization, Springer, Berlin, blz. 165-172. https://doi.org/10.1007/3-540-26981-9_20

Outlier identification rules for generalized linear models. / Kuhnt, S.; Pawlitschko, J.
Innovations in Classification, Data Science, and Information Systems (Proceedings of the 27th Annual Conference of the Gesellschaft für Klassifikation e.V., Cottbus, Germany, March 12–14, 2003), Part II. uitgave / D. Baier; K.D. Warnecke. Berlin: Springer, 2005. blz. 165-172 (Studies in Classification, Data Analysis, and Knowledge Organization).

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/Congresprocedure › Conferentiebijdrage › Academic › peer review

TY - GEN

T1 - Outlier identification rules for generalized linear models

AU - Kuhnt, S.

AU - Pawlitschko, J.

PY - 2005

Y1 - 2005

N2 - Observations which seem to deviate strongly from the main part of the data may occur in every statistical analysis. These observations, usually labelled as outliers, may cause completely misleading results when using standard methods and may also contain information about special events or dependencies. We discuss outliers in situations where a generalized linear model is assumed as null model for the regular data and introduce rules for their identification. For the special cases of a loglinear Poisson model and a logistic regression model some one-step identifiers based on robust and non-robust estimators are proposed and compared.

AB - Observations which seem to deviate strongly from the main part of the data may occur in every statistical analysis. These observations, usually labelled as outliers, may cause completely misleading results when using standard methods and may also contain information about special events or dependencies. We discuss outliers in situations where a generalized linear model is assumed as null model for the regular data and introduce rules for their identification. For the special cases of a loglinear Poisson model and a logistic regression model some one-step identifiers based on robust and non-robust estimators are proposed and compared.

U2 - 10.1007/3-540-26981-9_20

DO - 10.1007/3-540-26981-9_20

M3 - Conference contribution

SN - 3-540-23221-4

SN - 978-3-540-23221-6

T3 - Studies in Classification, Data Analysis, and Knowledge Organization

SP - 165

EP - 172

BT - Innovations in Classification, Data Science, and Information Systems (Proceedings of the 27th Annual Conference of the Gesellschaft für Klassifikation e.V., Cottbus, Germany, March 12–14, 2003), Part II

A2 - Baier, D.

A2 - Warnecke, K.D.

PB - Springer

CY - Berlin

ER -

Kuhnt S, Pawlitschko J. Outlier identification rules for generalized linear models. In Baier D, Warnecke KD, redacteurs, Innovations in Classification, Data Science, and Information Systems (Proceedings of the 27th Annual Conference of the Gesellschaft für Klassifikation e.V., Cottbus, Germany, March 12–14, 2003), Part II. Berlin: Springer. 2005. blz. 165-172. (Studies in Classification, Data Analysis, and Knowledge Organization). doi: 10.1007/3-540-26981-9_20

Outlier identification rules for generalized linear models

Samenvatting

Publicatie series

Toegang tot document

Vingerafdruk

Citeer dit