Initializing probabilistic linear discriminant analysis

Conference paper


Moschoglou, S., Nicolaou, M., Panagakis, Y. and Zafeiriou, S. 2017. Initializing probabilistic linear discriminant analysis. 2017 25th European Signal Processing Conference (EUSIPCO). Kos, Greece 28 Aug - 02 Sep 2017 Institute of Electrical and Electronics Engineers (IEEE). pp. 1175-1179 https://doi.org/10.23919/EUSIPCO.2017.8081393
TypeConference paper
TitleInitializing probabilistic linear discriminant analysis
AuthorsMoschoglou, S., Nicolaou, M., Panagakis, Y. and Zafeiriou, S.
Abstract

Component Analysis (CA) consists of a set of statistical techniques that decompose data to appropriate latent components that are relevant to the task-at-hand (e.g., clustering, segmentation, classification, alignment). During the past few years, an explosion of research in probabilistic CA has been witnessed, with the introduction of several novel methods (e.g., Probabilistic Principal Component Analysis, Probabilistic Linear Discriminant Analysis (PLDA), Probabilistic Canonical Correlation Analysis). PLDA constitutes one of the most widely used supervised CA techniques which is utilized in order to extract suitable, distinct subspaces by exploiting the knowledge of data annotated in terms of different labels. Nevertheless, an inherent difficulty in PLDA variants is the proper initialization of the parameters in order to avoid ending up in poor local maxima. In this light, we propose a novel method to initialize the parameters in PLDA in a consistent and robust way. The performance of the algorithm is demonstrated via a set of experiments on the modified XM2VTS database, which is provided by the authors of the original PLDA model.

LanguageEnglish
Conference2017 25th European Signal Processing Conference (EUSIPCO)
Page range1175-1179
ISSN2076-1465
ISBN
Hardcover9780992862671
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Publication dates
Print26 Oct 2017
Publication process dates
Deposited06 Mar 2018
Accepted25 May 2017
Output statusPublished
Digital Object Identifier (DOI)https://doi.org/10.23919/EUSIPCO.2017.8081393
Book title2017 25th European Signal Processing Conference (EUSIPCO)
Permalink -

https://repository.mdx.ac.uk/item/87878

  • 9
    total views
  • 0
    total downloads
  • 1
    views this month
  • 0
    downloads this month

Export as