previous | colloquium schedule | next

Colloquium - 19 Jan 1999

Department of Statistics
Texas A&M University

ABSTRACT: Some computer intensive methods such as nonparametric smoothing and bootstrap are extended to multiparameter models. For example, consider a likelihood model for clustered binary data. Besides a parameter for the success probability (which depends on some covariate, e.g. dose), there also are one or more correlation parameters in the model. We define local polynomial estimators for both mean and correlation as a function of the covariate. It can be shown that the estimators are strongly consistent and that they have a joint asymptotically normal distribution. As an alternative to this asymptotic distribution, we propose a bootstrap approach based on a one-step estimator, which can be used to construct simultaneous confidence intervals on a finite grid. For the multiple covariate case, we focus attention to statistical properties of local polynomial estimators in additive models.

Refreshments will be served in the Blocker Building, Room 447, at 3:30 p.m.

Copyright © Unless otherwise stated, all contents of these web pages are copyright by their respective authors or by the Department of Statistics, Texas A&M University, 1995-1998. Comments and questions about this site should be sent to web@stat.tamu.edu.