Coordinating principal component analyzers

Abstract : Mixtures of Principal Component Analyzers can be used to model high dimensional data that lie on or near a low dimensional manifold. By linearly mapping the PCA subspaces to one global low dimensional space, we obtain a 'global' low dimensional coordinate sys- tem for the data. As shown by Roweis et al., ensuring consistent global low-dimensional coordinates for the data can be expressed as a penal- ized likelihood optimization problem. We show that a restricted form of the Mixtures of Probabilistic PCA model allows for a more efficient algorithm. Experimental results are provided to illustrate the viability method.
Document type :
Conference papers
Complete list of metadatas

Cited literature [9 references]  Display  Hide  Download


https://hal.inria.fr/inria-00321498
Contributor : Jakob Verbeek <>
Submitted on : Wednesday, February 16, 2011 - 5:13:34 PM
Last modification on : Monday, September 25, 2017 - 10:08:04 AM
Long-term archiving on : Tuesday, May 17, 2011 - 2:31:03 AM

Files

verbeek02icann.pdf
Files produced by the author(s)

Identifiers

Citation

Jakob Verbeek, Nikos Vlassis, Ben Krose. Coordinating principal component analyzers. International Conference on Artificial Neural Networks (ICANN '02), Aug 2002, Madrid, Spain. ⟨10.1007/3-540-46084-5_148⟩. ⟨inria-00321498⟩

Share

Metrics

Record views

194

Files downloads

412