Orthogonal neighborhood preserving projections: A projection-based dimensionality reduction technique

Effrosyni Kokiopoulou, Yousef Saad

Research output: Contribution to journalArticlepeer-review

255 Scopus citations


This paper considers the problem of dimensionality reduction by orthogonal projection techniques. The main feature of the proposed techniques is that they attempt to preserve both the intrinsic neighborhood geometry of the data samples and the global geometry. In particular we propose a method, named Orthogonal Neighborhood Preserving Projections, which works by first building an "affinity" graph for the data, in a way that is similar to the method of Locally Linear Embedding (LLE). However, in contrast with the standard LLE where the mapping between the input and the reduced spaces is implicit, ONPP employs an explicit linear mapping between the two. As a result, handling new data samples becomes straightforward, as this amounts to a simple linear transformation.We show how to define kernel variants of ONPP, as well as how to apply the method in a supervised setting. Numerical experiments are reported to illustrate the performance of ONPP and to compare it with a few competing methods.

Original languageEnglish (US)
Pages (from-to)2143-2156
Number of pages14
JournalIEEE Transactions on Pattern Analysis and Machine Intelligence
Issue number12
StatePublished - Dec 2007

Bibliographical note

Funding Information:
The authors are grateful to Professor D. Boley for his valuable help and insightful discussions on various aspects of the paper. This work was supported by the US National Science Foundation under grant DMS 0510131 and by the Minnesota Supercomputing Institute.


  • Data visualization
  • Face recognition
  • Linear dimensionality reduction

Fingerprint Dive into the research topics of 'Orthogonal neighborhood preserving projections: A projection-based dimensionality reduction technique'. Together they form a unique fingerprint.

Cite this