The Resource Model selection criteria based on Kullback information measures for Weibull, logistic, and nonlinear regression frameworks, by Hyun-Joo Kim

Model selection criteria based on Kullback information measures for Weibull, logistic, and nonlinear regression frameworks, by Hyun-Joo Kim

Label
Model selection criteria based on Kullback information measures for Weibull, logistic, and nonlinear regression frameworks
Title
Model selection criteria based on Kullback information measures for Weibull, logistic, and nonlinear regression frameworks
Statement of responsibility
by Hyun-Joo Kim
Title variation
Model selection based on Kullback information
Creator
Subject
Language
eng
Summary
In statistical modeling, selecting an optimal model from a class of candidates is a critical issue. Over the past three decades, a number of model selection criteria have been proposed based on estimating Kullback's (1968) directed divergence between the model generating the data and a fitted candidate model. The Akaike (1973, 1974) information criterion, AIC, was the first such estimator. AIC is justified in a very general framework, and as a result, offers a crude estimator of the directed divergence: one which exhibits a potentially high degree of negative bias in small-sample applications (Hurvich and Tsai, 1989). The corrected Akaike information criterion (Hurvich and Tsai, 1989), AICc, adjusts for this bias, and as a result, often dramatically outperforms AIC as a model selection criterion. However, AICc is less broadly applicable than AIC, since its justification depends upon the form of the candidate model. AIC I (Hurvich, Shumway, and Tsai, 1990) is an improved version of AICc featuring a simulated bias correction. Recently, Kullback's (1968) symmetric divergence has been proposed for model selection as an alternate gauge of the separation between the generating model and the fitted approximating model (Cavanaugh, 1999, 2000). KIC, KICc, and KIC I are respective analogues of AIC, AICc, and AIC I based on the symmetric divergence (Cavanaugh, 2000). In this thesis, we develop AICc, AIC I , KICc, and KIC I based on the quasi-likelihood, and AIC I and KIC I based on the ordinary likelihood, for the frameworks of Weibull regression, traditional logistic regression, and logistic regression with extra variation. Also, we develop AICc, AIC I , KICc, and KIC I based on the ordinary likelihood for the framework of nonlinear regression. We evaluate the selection performance of AIC, AICc, AIC I , KIC, KICc, and KIC I in simulation studies
Additional physical form
Also available on the Internet.
Cataloging source
MUU
http://library.link/vocab/creatorDate
1971-
http://library.link/vocab/creatorName
Kim, Hyun-Joo
Degree
Ph. D.
Dissertation year
2000.
Government publication
government publication of a state province territory dependency etc
Granting institution
University of Missouri-Columbia
Illustrations
illustrations
Index
no index present
Literary form
non fiction
Nature of contents
  • bibliography
  • theses
http://library.link/vocab/subjectName
  • Weibull distribution
  • Logistic distribution
  • Regression analysis
  • Estimation theory
  • Mathematical models
Target audience
specialized
Label
Model selection criteria based on Kullback information measures for Weibull, logistic, and nonlinear regression frameworks, by Hyun-Joo Kim
Instantiates
Publication
Note
  • Typescript
  • Vita
Bibliography note
Includes bibliographical references (leaves 104-107)
Carrier category
online resource
Carrier category code
  • cr
Carrier MARC source
rdacarrier
Content category
text
Content type code
  • txt
Content type MARC source
rdacontent
Control code
50048229
Dimensions
29 cm
Dimensions
unknown
Extent
viii, 126 leaves
Media category
computer
Media MARC source
rdamedia
Media type code
  • c
Other physical details
illustrations
Specific material designation
remote
Label
Model selection criteria based on Kullback information measures for Weibull, logistic, and nonlinear regression frameworks, by Hyun-Joo Kim
Publication
Note
  • Typescript
  • Vita
Bibliography note
Includes bibliographical references (leaves 104-107)
Carrier category
online resource
Carrier category code
  • cr
Carrier MARC source
rdacarrier
Content category
text
Content type code
  • txt
Content type MARC source
rdacontent
Control code
50048229
Dimensions
29 cm
Dimensions
unknown
Extent
viii, 126 leaves
Media category
computer
Media MARC source
rdamedia
Media type code
  • c
Other physical details
illustrations
Specific material designation
remote

Library Locations

    • University of Missouri Libraries DepositoryBorrow it
      2908 Lemone Blvd, Columbia, MO, 65211, US
      38.919360 -92.291620
Processing Feedback ...