Latent Transport Models for Multivariate Functional Data

07/12/2021
by   Cody Carroll, et al.
0

Multivariate functional data present theoretical and practical complications which are not found in univariate functional data. One of these is a situation where the component functions of multivariate functional data are positive and are subject to mutual time warping. That is, the component processes exhibit a similar shape but are subject to systematic phase variation across their time domains. We introduce a novel model for multivariate functional data that incorporates such mutual time warping via nonlinear transport functions. This model allows for meaningful interpretation and is well suited to represent functional vector data. The proposed approach combines a random amplitude factor for each component with population based registration across the components of a multivariate functional data vector and also includes a latent population function, which corresponds to a common underlying trajectory as well as subject-specific warping component. We also propose estimators for all components of the model. The proposed approach not only leads to a novel representation for multivariate functional data, but is also useful for downstream analyses such as Fréchet regression. Rates of convergence are established when curves are fully observed or observed with measurement error. The usefulness of the model, interpretations and practical aspects are illustrated in simulations and with application to multivariate human growth curves as well as multivariate environmental pollution data.

READ FULL TEXT

page 7

page 14

page 18

research
11/04/2018

Cross-Component Registration for Multivariate Functional Data with Application to Longitudinal Growth Curves

Multivariate functional data are becoming ubiquitous with the advance of...
research
10/31/2018

A General Framework for Multivariate Functional Principal Component Analysis of Amplitude and Phase Variation

Functional data typically contains amplitude and phase variation. In man...
research
07/22/2022

Spatially Penalised Registration of Multivariate Functional Data

Registration of multivariate functional data involves handling of both c...
research
06/02/2021

Filtrated Common Functional Principal Components for Multivariate Functional data

Local field potentials (LFPs) are signals that measure electrical activi...
research
10/15/2018

Rank Dynamics for Functional Data

We study the dynamic behavior of cross-sectional ranks over time for fun...
research
06/24/2020

Uniform convergence of local Fréchet regression and time warping for metric-space-valued trajectories

For real-valued functional data, it is well known that failure to separa...
research
04/09/2017

An Outlyingness Matrix for Multivariate Functional Data Classification

The classification of multivariate functional data is an important task ...

Please sign up or login with your details

Forgot password? Click here to reset