Guest Speaker: Ana-Maria Staicu, Dept. of Statistics, N.C. State University
Abstract: When data are observed at high frequency or the repeated measures exhibit a correlation that cannot be described accurately by parametric models, functional data techniques are often used.
I will discuss smoothing using pre-specified basis functions and functional principal component analysis as common approaches to model functional data. If time permits we will consider dependent functional data that are correlated because of a longitudinal-based design: each subject is observed at repeated times and at each time a functional observation (e.g. curve, image) is recorded. The methods will be illustrated numerically through data applications.
To view this presentation, please CLICK HERE