prev | colloquium schedule | next

Colloquium - 4 May 2000

Department of Statistics
Texas A&M University

ABSTRACT: Data in the form of functions or curves are increasingly common in fields such as biology, genetics, geosciences and finance. Some longitudinal data can also be cast in this framework. We assume that per subject or unit we observe the realization of a stochastic process, a random response function, in addition to covariates which can be finite-dimensional or of functional type themselves.

One is then confronted with a non-standard regression problem where the response is a random function. Several approaches to modeling regression for such situations will be discussed. These include the functional linear model and a new proposal which is based on the eigenfunction decomposition of the response function. For the latter situation, we propose to regress each principal component of the random response function on the set of predictors. Thus the problem is broken down into a series of classical regression problems with possibly high-dimensional predictors.

We aim to address these classical regression problems without having to specify a fully parametric model, while still escaping the curse of dimension. This goal is achieved with a class of single index models which have been termed QLUEs for quasi-likelihood with nonparametric link and variance function estimates. The proposed methods are illustrated with data on reproduction and lifespan of medflies (Co-authors on various parts include J. Chiou, G. He, J.L. Wang).

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, 1995-2001. Comments and questions about this site should be sent to web@stat.tamu.edu.