Title: Factor analysis and nuisance attribute projection revisited
Other Titles: Nový pohled na faktorovou analýza a metodu Nuisance Attribute Projection
Authors: Machlica, Lukáš
Zajíc, Zbyněk
Citation: MACHLICA, Lukáš; ZAJÍC, Zbyněk. Factor analysis and nuisance attribute projection revisited. In: Proceedings of ICSPL 2012: 13th Annual Conference of the International Speech Communication Association 2012, Portland, Oregon, USA, 9-13 September 2012. [Baixas]: ISCA, 2012, p. 1570-1573. IBSN 978-1-62276-759-5.
Issue Date: 2012
Publisher: ISCA
Document type: článek
article
URI: http://www.kky.zcu.cz/cs/publications/LukasMachlica_2012_FactorAnalysisand
http://hdl.handle.net/11025/17039
ISBN: 978-1-62276-759-5
Keywords: NAP;FA;PCA;nejmenší ÄŤtverce
Keywords in different language: NAP;FA;PCA;least squares
Abstract: V článku jsou popsány, analyzovány a porovnány metoda faktorové analýzy a metoda projekce nežádoucích atributů (Nuisance Attribute Projection - NAP). Je ukázáno jak je obě metody možné vyjádřit a propojit využitím metody nejmenších čtverců.
Abstract in different language: In the paper Factor Analysis (FA) and Nuisance Attribute Projection (NAP) are reviewed, analyzed and compared. Since nowadays FA become a part of most state-of-the-art recognition systems (used e.g. in the concept of i-vectors or PLDA models) it is of relevance to examine different insights into the problem. NAP was chosen as a counterpart to FA as an advanced PCA like method often utilized in speaker recognition systems along with FA. It is demonstrated how can be both FA and NAP expressed as solutions of Least Squares (LS), the consequences of the LS formulation are discussed, and it is shown in what extent do solutions of NAP and FA overlap.
Rights: © Lukáš Machlica - Zbyněk Zajíc
Appears in Collections:Články / Articles (KKY)
Články / Articles (NTIS)

Files in This Item:
File Description SizeFormat 
LukasMachlica_2012_FactorAnalysisand.pdfPlný text180,99 kBAdobe PDFView/Open


Please use this identifier to cite or link to this item: http://hdl.handle.net/11025/17039

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.