Advanced search
1 file | 146.94 KB

Imprecise probability models for inference in exponential families

Erik Quaeghebeur (UGent) and Gert De Cooman (UGent)
Author
Organization
Abstract
When considering sampling models described by a distribution from an exponential family, it is possible to create two types of imprecise probability models. One is based on the corresponding conjugate distribution and the other on the corresponding predictive distribution. In this paper, we show how these types of models can be constructed for any (regular, linear, canonical) exponential family, such as the centered normal distribution. To illustrate the possible use of such models, we take a look at credal classification. We show that they are very natural and potentially promising candidates for describing the attributes of a credal classifier, also in the case of continuous attributes.
Keywords
Exponential family, Inference, Imprecise probability models, Conjugate analysis, Naive credal classifier

Downloads

  • EQ-2005-ISIPTA-paper.pdf
    • full text
    • |
    • open access
    • |
    • PDF
    • |
    • 146.94 KB

Citation

Please use this url to cite or link to this publication:

Chicago
Quaeghebeur, Erik, and Gert De Cooman. 2005. “Imprecise Probability Models for Inference in Exponential Families.” In ISIPTA 05-PROCEEDINGS OF THE FOURTH INTERNATIONAL SYMPOSIUM ON IMPRECISE PROBABILITIES AND THEIR APPLICATIONS, ed. Fabio G Cozman, Robert Nau, and Teddy Seidenfeld, 287–296. Manno, Switzerland: International Society for Imprecise Probability: Theories and Applications (SIPTA).
APA
Quaeghebeur, E., & De Cooman, G. (2005). Imprecise probability models for inference in exponential families. In F. G. Cozman, R. Nau, & T. Seidenfeld (Eds.), ISIPTA 05-PROCEEDINGS OF THE FOURTH INTERNATIONAL SYMPOSIUM ON IMPRECISE PROBABILITIES AND THEIR APPLICATIONS (pp. 287–296). Presented at the 4th International Symposium on Imprecise Probabilities and Their Applications, Manno, Switzerland: International Society for Imprecise Probability: Theories and Applications (SIPTA).
Vancouver
1.
Quaeghebeur E, De Cooman G. Imprecise probability models for inference in exponential families. In: Cozman FG, Nau R, Seidenfeld T, editors. ISIPTA 05-PROCEEDINGS OF THE FOURTH INTERNATIONAL SYMPOSIUM ON IMPRECISE PROBABILITIES AND THEIR APPLICATIONS. Manno, Switzerland: International Society for Imprecise Probability: Theories and Applications (SIPTA); 2005. p. 287–96.
MLA
Quaeghebeur, Erik, and Gert De Cooman. “Imprecise Probability Models for Inference in Exponential Families.” ISIPTA 05-PROCEEDINGS OF THE FOURTH INTERNATIONAL SYMPOSIUM ON IMPRECISE PROBABILITIES AND THEIR APPLICATIONS. Ed. Fabio G Cozman, Robert Nau, & Teddy Seidenfeld. Manno, Switzerland: International Society for Imprecise Probability: Theories and Applications (SIPTA), 2005. 287–296. Print.
@inproceedings{320871,
  abstract     = {When considering sampling models described by a distribution from an exponential family, it is possible to create two types of imprecise probability models. One is based on the corresponding conjugate distribution and the other on the corresponding predictive distribution. In this paper, we show how these types of models can be constructed for any (regular, linear, canonical) exponential family, such as the centered normal distribution.
To illustrate the possible use of such models, we take a look at credal classification. We show that they are very natural and potentially promising candidates for describing the attributes of a credal classifier, also in the case of continuous attributes.},
  author       = {Quaeghebeur, Erik and De Cooman, Gert},
  booktitle    = {ISIPTA 05-PROCEEDINGS OF THE FOURTH INTERNATIONAL SYMPOSIUM ON IMPRECISE PROBABILITIES AND THEIR APPLICATIONS},
  editor       = {Cozman, Fabio G and Nau, Robert and Seidenfeld, Teddy},
  keyword      = {Exponential family,Inference,Imprecise probability models,Conjugate analysis,Naive credal classifier},
  language     = {eng},
  location     = {Pittsburgh, PA, USA},
  pages        = {287--296},
  publisher    = {International Society for Imprecise Probability: Theories and Applications (SIPTA)},
  title        = {Imprecise probability models for inference in exponential families},
  url          = {http://www.sipta.org/isipta05/proceedings/019.html},
  year         = {2005},
}

Web of Science
Times cited: