American Statistical Association
New York City
Metropolitan Area Chapter

Mailman School of Public Health
Columbia University
Department of Epidemiology Colloquium



Sara Lopez Pintado
Department of Statistics
Rutgers University


The statistical analysis of functional data (or curves) is a growing need in many research areas such as medicine, biology, or engineering. Many multivariate methods (e.g. analysis of variance and classification) have been extended to functional data. The statistical analysis of curves can be much improved by using robust estimators. In the talk, new definitions of depth for functional data, which provide a way of ordering curves from center-outward, will be discussed. Robust estimates such as median curve, trimmed mean, trimmed regions, contours and central regions are extended to functions and their structural properties are studied. Simulation results show that the location estimates based on the new notions of depth present a better behavior than the mean or coordinate-wise trimmed mean for contaminated models. A scale curve is introduced to describe dispersion in a sample of functions. These inferential methods are applied to various real data sets. We have also generalized the Wilcoxon rank sum test to functions. This allows us to test whether two groups of curves come from the same population. Application of these tools with children growth curves will be discussed.

Date: Monday, April 30, 2007
Time: 1:00 - 2:30 P.M.
Location: Mailman School of Public Health
Department of Epidemiology
722 West 168th Street
9th Floor Conference Room
New York, New York

Home Page | Chapter News | Chapter Officers | Chapter Events
Other Metro Area Events | ASA National Home Page | Links To Other Websites

Copyright © 2007 by New York City Metropolitan Area Chapter of the ASA
Designed and maintained by Cynthia Scherer
Send questions or comments to