A penalization method to estimate the intrinsic dimensionality of data

We propose a novel penalization method for estimating the intrinsic dimensionality of data within a Probabilistic Principal Components Model, extending beyond the Gaussian case. Unlike existing approaches, our method is designed to handle non-normal data, providing a flexible alternative to traditio...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Forzani, Liliana, Rodriguez, Daniela, Sued, Mariela
Formato: Artículo
Lenguaje:Inglés
Publicado: Statistical Papers (e-ISSN: 1613-9798) 2025
Materias:
Análisis de Datos
Data Analysis
Estadística
Statistics
Acceso en línea:https://repositorio.utdt.edu/handle/20.500.13098/13449
Aporte de:
Repositorio Digital Universidad Torcuato Di Tella de Universidad Torcuato Di Tella
id I57-R163-20.500.13098-13449
record_format dspace
spelling I57-R163-20.500.13098-134492025-06-07T05:03:00Z A penalization method to estimate the intrinsic dimensionality of data Forzani, Liliana Rodriguez, Daniela Sued, Mariela Análisis de Datos Data Analysis Estadística Statistics We propose a novel penalization method for estimating the intrinsic dimensionality of data within a Probabilistic Principal Components Model, extending beyond the Gaussian case. Unlike existing approaches, our method is designed to handle non-normal data, providing a flexible alternative to traditional factor models. Our procedure identifies the dimension at which the eigenvalues of a scatter matrix stabilize. We establish the consistency of the procedure under mild conditions and demonstrate its robustness across a range of data distributions. A comparative analysis highlights its advantages over existing techniques, making it a valuable tool for dimensionality estimation without relying on distributional assumptions. Forzani, L., Rodriguez, D. & Sued, M. A penalization method to estimate the intrinsic dimensionality of data. Stat Papers 66, 46 (2025). https://doi.org/10.1007/s00362-025-01667-0 Statistical Papers (e-ISSN: 1613-9798) 2025-06-06T20:39:28Z 2025-02-06 info:eu-repo/semantics/article https://repositorio.utdt.edu/handle/20.500.13098/13449 eng Statistical Papers (e-ISSN: 1613-9798), Volume 66, article number 46 info:eu-repo/semantics/restrictedAccess http://rightsstatements.org/page/InC/1.0/?language=es 20 p. application/pdf application/pdf
institution Universidad Torcuato Di Tella
institution_str I-57
repository_str R-163
collection Repositorio Digital Universidad Torcuato Di Tella
language Inglés
orig_language_str_mv eng
topic Análisis de Datos
Data Analysis
Estadística
Statistics
spellingShingle Análisis de Datos
Data Analysis
Estadística
Statistics
Forzani, Liliana
Rodriguez, Daniela
Sued, Mariela
A penalization method to estimate the intrinsic dimensionality of data
topic_facet Análisis de Datos
Data Analysis
Estadística
Statistics
description We propose a novel penalization method for estimating the intrinsic dimensionality of data within a Probabilistic Principal Components Model, extending beyond the Gaussian case. Unlike existing approaches, our method is designed to handle non-normal data, providing a flexible alternative to traditional factor models. Our procedure identifies the dimension at which the eigenvalues of a scatter matrix stabilize. We establish the consistency of the procedure under mild conditions and demonstrate its robustness across a range of data distributions. A comparative analysis highlights its advantages over existing techniques, making it a valuable tool for dimensionality estimation without relying on distributional assumptions.
format Artículo
author Forzani, Liliana
Rodriguez, Daniela
Sued, Mariela
author_facet Forzani, Liliana
Rodriguez, Daniela
Sued, Mariela
author_sort Forzani, Liliana
title A penalization method to estimate the intrinsic dimensionality of data
title_short A penalization method to estimate the intrinsic dimensionality of data
title_full A penalization method to estimate the intrinsic dimensionality of data
title_fullStr A penalization method to estimate the intrinsic dimensionality of data
title_full_unstemmed A penalization method to estimate the intrinsic dimensionality of data
title_sort penalization method to estimate the intrinsic dimensionality of data
publisher Statistical Papers (e-ISSN: 1613-9798)
publishDate 2025
url https://repositorio.utdt.edu/handle/20.500.13098/13449
work_keys_str_mv AT forzanililiana apenalizationmethodtoestimatetheintrinsicdimensionalityofdata
AT rodriguezdaniela apenalizationmethodtoestimatetheintrinsicdimensionalityofdata
AT suedmariela apenalizationmethodtoestimatetheintrinsicdimensionalityofdata
AT forzanililiana penalizationmethodtoestimatetheintrinsicdimensionalityofdata
AT rodriguezdaniela penalizationmethodtoestimatetheintrinsicdimensionalityofdata
AT suedmariela penalizationmethodtoestimatetheintrinsicdimensionalityofdata
_version_ 1842217744770007040

Ejemplares similares