In this article, we address the problem of mining and analyzing multivariate functional data. That is, data where each observation is a set of possibly correlated functions. Complex data of this kind is more and more common in many research fields, particularly in the biomedical context. In this work, we propose and apply a new concept of depth measure for multivariate functional data. With this new depth measure it is possible to generalize robust statistics, such as the median, to the multivariate functional framework, which in turn allows the application of outlier detection, boxplots construction, and nonparametric tests also in this more general framework. We present an application to Electrocardiographic (ECG) signals.

Depth Measures for Multivariate Functional Data / F. Ieva, A.M. Paganoni. - In: COMMUNICATIONS IN STATISTICS. THEORY AND METHODS. - ISSN 0361-0926. - 42:7(2013), pp. 1265-1276.

Depth Measures for Multivariate Functional Data

F. Ieva
Primo
;
2013

Abstract

In this article, we address the problem of mining and analyzing multivariate functional data. That is, data where each observation is a set of possibly correlated functions. Complex data of this kind is more and more common in many research fields, particularly in the biomedical context. In this work, we propose and apply a new concept of depth measure for multivariate functional data. With this new depth measure it is possible to generalize robust statistics, such as the median, to the multivariate functional framework, which in turn allows the application of outlier detection, boxplots construction, and nonparametric tests also in this more general framework. We present an application to Electrocardiographic (ECG) signals.
Depth measures ; ECG signals ; Multivariate functional data ; Rank tests
Settore SECS-S/01 - Statistica
Settore MED/01 - Statistica Medica
Article (author)
File in questo prodotto:
File Dimensione Formato  
CommStat DEF.pdf

accesso aperto

Tipologia: Publisher's version/PDF
Dimensione 554.86 kB
Formato Adobe PDF
554.86 kB Adobe PDF Visualizza/Apri
Pubblicazioni consigliate

Caricamento pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/2434/233394
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 34
  • ???jsp.display-item.citation.isi??? 32
social impact