UNIGE document Doctoral Thesis
previous document  unige:45953  next document
add to browser collection

Contributions to overdispersed count data modeling: robustness, small samples and other extensions

Defense Thèse de doctorat : Univ. Genève, 2015 - GSEM 3 - 2015/01/23
Abstract This PhD thesis contributes to the modeling of overdispered count data in three ways. First, we extend two approaches for building robust M-estimators of the regression parameters in the class of generalized linear models to the negative binomial (NB) distribution. Second, we adapt recently developed tests, such as the so-called saddlepoint test, to the framework of overdispersed count data and give a detailed account of their computation and implementation. Through extensive simulations we compare them to traditional tests in order to assess their effective level under the null hypothesis and their power under alternatives, and this for models based on a full NB likelihood or on moment restrictions only. Finally, we enlarge our scope to the class of mixed compound Poisson models, where the overdispersion is due to the variance of unobserved mixing variables. We propose a semi-parametric framework where the mixing distribution is left unspecified and estimated by point-masses.
Keywords Empirical LikelihoodExponential TiltingMixed Compound Poisson ModelNegative Binomial RegressionRobust M-EstimatorsSaddlepoint Test.
URN: urn:nbn:ch:unige-459536
Full text
Thesis (6.8 MB) - document accessible for UNIGE members only Limited access to UNIGE
Research group Probabilités et statistique
(ISO format)
AEBERHARD, William. Contributions to overdispersed count data modeling: robustness, small samples and other extensions. Université de Genève. Thèse, 2015. doi: 10.13097/archive-ouverte/unige:45953 https://archive-ouverte.unige.ch/unige:45953

763 hits



Deposited on : 2015-01-28

Export document
Format :
Citation style :