Dimensionality Reduction of Clustered Data Sets

G. Sanguinetti

doi:10.1109/TPAMI.2007.70819

Dimensionality Reduction of Clustered Data Sets

G. Sanguinetti

School of Informatics

Research output: Contribution to journal › Article › peer-review

Abstract

We present a novel probabilistic latent variable model to perform linear dimensionality reduction on data sets which contain clusters. We prove that the maximum likelihood solution of the model is an unsupervised generalization of linear discriminant analysis. This provides a completely new approach to one of the most established and widely used classification algorithms. The performance of the model is then demonstrated on a number of real and artificial data sets.

Original language	English
Pages (from-to)	535-540
Number of pages	6
Journal	IEEE Transactions on Pattern Analysis and Machine Intelligence
Volume	30
Issue number	3
DOIs	https://doi.org/10.1109/TPAMI.2007.70819
Publication status	Published - Mar 2008

Access to Document

10.1109/TPAMI.2007.70819

http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=4378396

Cite this

@article{9c755639d55b4aa2814c66f7ff277208,

title = "Dimensionality Reduction of Clustered Data Sets",

abstract = "We present a novel probabilistic latent variable model to perform linear dimensionality reduction on data sets which contain clusters. We prove that the maximum likelihood solution of the model is an unsupervised generalization of linear discriminant analysis. This provides a completely new approach to one of the most established and widely used classification algorithms. The performance of the model is then demonstrated on a number of real and artificial data sets.",

author = "G. Sanguinetti",

year = "2008",

month = mar,

doi = "10.1109/TPAMI.2007.70819",

language = "English",

volume = "30",

pages = "535--540",

journal = "IEEE Transactions on Pattern Analysis and Machine Intelligence",

issn = "0162-8828",

publisher = "Institute of Electrical and Electronics Engineers",

number = "3",

}

TY - JOUR

T1 - Dimensionality Reduction of Clustered Data Sets

AU - Sanguinetti, G.

PY - 2008/3

Y1 - 2008/3

N2 - We present a novel probabilistic latent variable model to perform linear dimensionality reduction on data sets which contain clusters. We prove that the maximum likelihood solution of the model is an unsupervised generalization of linear discriminant analysis. This provides a completely new approach to one of the most established and widely used classification algorithms. The performance of the model is then demonstrated on a number of real and artificial data sets.

AB - We present a novel probabilistic latent variable model to perform linear dimensionality reduction on data sets which contain clusters. We prove that the maximum likelihood solution of the model is an unsupervised generalization of linear discriminant analysis. This provides a completely new approach to one of the most established and widely used classification algorithms. The performance of the model is then demonstrated on a number of real and artificial data sets.

U2 - 10.1109/TPAMI.2007.70819

DO - 10.1109/TPAMI.2007.70819

M3 - Article

SN - 0162-8828

VL - 30

SP - 535

EP - 540

JO - IEEE Transactions on Pattern Analysis and Machine Intelligence

JF - IEEE Transactions on Pattern Analysis and Machine Intelligence

IS - 3

ER -

Dimensionality Reduction of Clustered Data Sets

Abstract

Access to Document

Fingerprint

Cite this