A0724
Title: Analysis of functional data: From serial to spatial dependence
Authors: Pipat Wongsa-art - Cardiff University (United Kingdom) [presenting]
Abstract: Today, a widely accepted description of functional data is ``multivariate data with an ordering on the dimensions''. Such a definition suggests that this type of data consists of curves that vary over a continuum of e.g., time, frequency, or wavelength. In economics and finance, we often must deal with functional data that are either temporally or spatially dependent. We first discuss a novel method that can help to capture the temporal dependence (i.e., dynamics) of functional time series, e.g., time series of correlation functions that describe co-movements of financial returns. We then explain how such a method can be extended to obtain a new method for analysing functional data under spatial dependence. We illustrate the usefulness of this tool with simulation exercises.
