A generalized Mahalanobis distance for the classification of functional data

Andrea Ghiglietti; Francesca Ieva; Anna Maria Paganoni

A generalized Mahalanobis distance for the classification of functional data

Andrea Ghiglietti, Francesca Ieva, Anna Maria Paganoni

Risultato della ricerca: Contributo in libro › Contributo a convegno

Abstract

We present some asymptotic results on the distance between the means of samples of curves generated by independent continuous time stochastic processes in L2(T). The asymptotic results are based on mild assumptions on the moments of the processes, and there are no conditions on their probability distribution. The metrics we consider extends the Mahalanobis distance to L2(T) without any truncation on the first principal components. Applications in the context of classification of functional data are finally discussed.

Lingua originale	English
Titolo della pubblicazione ospite	Classification and Data Analysis Group: Book of abstracts
Pagine	1-6
Numero di pagine	6
Stato di pubblicazione	Pubblicato - 2017
Pubblicato esternamente	Sì
Evento	Classification and Data Analysis Group 2017 - Milano Durata: 13 set 2017 → 15 set 2017

Convegno

Convegno	Classification and Data Analysis Group 2017
Città	Milano
Periodo	13/9/17 → 15/9/17

Keywords

Distances in L2
Functiona Data
Two-sample problems

Altri file e collegamenti

Link a IRIS PubliCatt

Cita questo

@inproceedings{4cf5ebffb09048d1a593113b02279c1c,

title = "A generalized Mahalanobis distance for the classification of functional data",

abstract = "We present some asymptotic results on the distance between the means of samples of curves generated by independent continuous time stochastic processes in L2(T). The asymptotic results are based on mild assumptions on the moments of the processes, and there are no conditions on their probability distribution. The metrics we consider extends the Mahalanobis distance to L2(T) without any truncation on the first principal components. Applications in the context of classification of functional data are finally discussed.",

keywords = "Distances in L2, Functiona Data, Two-sample problems, Distances in L2, Functiona Data, Two-sample problems",

author = "Andrea Ghiglietti and Francesca Ieva and Paganoni, {Anna Maria}",

year = "2017",

language = "English",

isbn = "978-88-99459-71-0",

pages = "1--6",

booktitle = "Classification and Data Analysis Group: Book of abstracts",

note = "Classification and Data Analysis Group 2017 ; Conference date: 13-09-2017 Through 15-09-2017",

}

TY - GEN

T1 - A generalized Mahalanobis distance for the classification of functional data

AU - Ghiglietti, Andrea

AU - Ieva, Francesca

AU - Paganoni, Anna Maria

PY - 2017

Y1 - 2017

N2 - We present some asymptotic results on the distance between the means of samples of curves generated by independent continuous time stochastic processes in L2(T). The asymptotic results are based on mild assumptions on the moments of the processes, and there are no conditions on their probability distribution. The metrics we consider extends the Mahalanobis distance to L2(T) without any truncation on the first principal components. Applications in the context of classification of functional data are finally discussed.

AB - We present some asymptotic results on the distance between the means of samples of curves generated by independent continuous time stochastic processes in L2(T). The asymptotic results are based on mild assumptions on the moments of the processes, and there are no conditions on their probability distribution. The metrics we consider extends the Mahalanobis distance to L2(T) without any truncation on the first principal components. Applications in the context of classification of functional data are finally discussed.

KW - Distances in L2

KW - Functiona Data

KW - Two-sample problems

KW - Distances in L2

KW - Functiona Data

KW - Two-sample problems

UR - http://hdl.handle.net/10807/117863

M3 - Conference contribution

SN - 978-88-99459-71-0

SP - 1

EP - 6

BT - Classification and Data Analysis Group: Book of abstracts

T2 - Classification and Data Analysis Group 2017

Y2 - 13 September 2017 through 15 September 2017

ER -

A generalized Mahalanobis distance for the classification of functional data

Abstract

Convegno

Keywords

Altri file e collegamenti

Fingerprint

Cita questo