ABSTRACT:

As science develops, functional data whose observations are functions rather than numbers or vectors are more and more easily collected in many fields such as biology, medicine sciences among others. Of interest is how to deal with such a kind of data. In this presentation, an overview of functional data analysis is given. Methodologies for functional hypothesis testing and automatic choice of dimensionality of principal components are proposed and studied. Real data applications and simulation studies conducted show that these methodologies are useful.

July 12, 1999