Topics in Models With Random Effects

Vortrag von Dr. Daniel Flores Agreda

Sprecher eingeladen von: Prof. Dr. Reinhard Furrer

Datum: 15.12.17  Zeit: 12.30 - 13.30  Raum: Y27H46

In the first part of this talk, the problem of Bootstrap inference for the parameters of a GLMM is addressed. We formulate a bootstrapping strategy consisting on the random weighting of the contributions to the Joint Likelihood of Outcomes and Random Effects. Using the Laplace Approximation method for integrals on this function, yields a Random Weighted Log-Likelihood that produces the desired bootstrap replicates after optimisation. In order to assess the properties of this procedure, that we name Random Weighted Likelihood Bootstrap (RWLB), we compare analytically their resulting EE to those of the Generalized Cluster Bootstrap for Gaussian LMM and conduct simulation studies both in a LMM and Mixed Logit regression contexts. As a potential application, this new scheme is adapted to the estimation of the uncertainty in prediction of random effects in a GLMM, as measured by the Mean Squared Error for the Predictors (MSEP). In the second part of this talk, we shall discuss the framework of Accelerated Failure Time models, an integration of random effects to the analysis of durations with interval-censored survival data and prospective area of study.