On the Mahalanobis Distance Classification Criterion for Multidimensional Normal Distributions

Gallego Bonet, Guillermo; Cuevas Rodríguez, Carlos; Mohedano del Pozo, Raúl y García Santos, Narciso (2013). On the Mahalanobis Distance Classification Criterion for Multidimensional Normal Distributions. "IEEE Transactions on Signal Processing", v. 61 (n. 17); pp. 4387-4396. ISSN 1053-587X. https://doi.org/10.1109/TSP.2013.2269047.

Descripción

Título: On the Mahalanobis Distance Classification Criterion for Multidimensional Normal Distributions
Autor/es:
  • Gallego Bonet, Guillermo
  • Cuevas Rodríguez, Carlos
  • Mohedano del Pozo, Raúl
  • García Santos, Narciso
Tipo de Documento: Artículo
Título de Revista/Publicación: IEEE Transactions on Signal Processing
Fecha: 1 Septiembre 2013
Volumen: 61
Materias:
Palabras Clave Informales: multidimensional signal processing, uncertainty, classification algorithms, Gaussian distribution, Chi-squared distribution, Mahalanobis distance.
Escuela: E.T.S.I. Telecomunicación (UPM)
Departamento: Señales, Sistemas y Radiocomunicaciones
Licencias Creative Commons: Reconocimiento - Sin obra derivada - No comercial

Texto completo

[img]
Vista Previa
PDF (Document Portable Format) - Se necesita un visor de ficheros PDF, como GSview, Xpdf o Adobe Acrobat Reader
Descargar (1MB) | Vista Previa

Resumen

Many existing engineering works model the statistical characteristics of the entities under study as normal distributions. These models are eventually used for decision making, requiring in practice the definition of the classification region corresponding to the desired confidence level. Surprisingly enough, however, a great amount of computer vision works using multidimensional normal models leave unspecified or fail to establish correct confidence regions due to misconceptions on the features of Gaussian functions or to wrong analogies with the unidimensional case. The resulting regions incur in deviations that can be unacceptable in high-dimensional models. Here we provide a comprehensive derivation of the optimal confidence regions for multivariate normal distributions of arbitrary dimensionality. To this end, firstly we derive the condition for region optimality of general continuous multidimensional distributions, and then we apply it to the widespread case of the normal probability density function. The obtained results are used to analyze the confidence error incurred by previous works related to vision research, showing that deviations caused by wrong regions may turn into unacceptable as dimensionality increases. To support the theoretical analysis, a quantitative example in the context of moving object detection by means of background modeling is given.

Más información

ID de Registro: 19074
Identificador DC: http://oa.upm.es/19074/
Identificador OAI: oai:oa.upm.es:19074
Identificador DOI: 10.1109/TSP.2013.2269047
Depositado por: Dr Guillermo Gallego Bonet
Depositado el: 28 Ago 2013 08:31
Ultima Modificación: 21 Abr 2016 17:19
  • Open Access
  • Open Access
  • Sherpa-Romeo
    Compruebe si la revista anglosajona en la que ha publicado un artículo permite también su publicación en abierto.
  • Dulcinea
    Compruebe si la revista española en la que ha publicado un artículo permite también su publicación en abierto.
  • Recolecta
  • e-ciencia
  • Observatorio I+D+i UPM
  • OpenCourseWare UPM