Course Title: Functional Data Analysis
Spring Semester 2002
Taught By: J. S. Marron
Description: Functional Data Analysis is an emerging
subfield of statistics. A way to understand the field is
to think about the "atom" of a statistical analysis. In a
first course in statistics, atoms are numbers, and one
analyzes the structure of populations of numbers. In
multivariate analysis, atoms are vectors. In Functional
Data Analysis, the atoms are more complex objects, e.g.
curves (this case is closely related to "longitudinal data
analysis"), images, and even shapes of objects in 3d.
Understanding and using a set of tools for analyzing such
data is the focus of the course. Methods discussed will
include related ideas from the old field of "pattern
recognition" and the rather new area called "machine
learning". Driving examples come from medical imaging
(where there is interest in a population analysis of
various body parts), and from genetic micro-array analysis.
Text Book:
required: none
recommended: Ramsay, J. O. and Silverman, B. W. (1997)
Functional Data Analysis, Springer.
Prerequisites: One year of probability and statistics, at
the undergraduate level.