Outlier detection in the framework of dimensionality reduction

Qiang Ye, Weifeng Zhi

Research output: Contribution to journalArticlepeer-review

7 Scopus citations


We propose an effective outlier detection algorithm for high-dimensional data. We consider manifold models of data as is typically assumed in dimensionality reduction/manifold learning. Namely, we consider a noisy data set sampled from a low-dimensional manifold in a high-dimensional data space. Our algorithm uses local geometric structure to determine inliers, from which the outliers are identified. The algorithm is applicable to both linear and nonlinear models of data. We also discuss various implementation issues and we present several examples to demonstrate the effectiveness of the new approach.

Original languageEnglish
Article number1550017
JournalInternational Journal of Pattern Recognition and Artificial Intelligence
Issue number4
StatePublished - Jun 28 2015

Bibliographical note

Funding Information:
We would like to thank three anonymous referees for their many valuable comments. This research was supported in part by National Science Foundation under grant DMS-1317424.

Publisher Copyright:
© 2015 World Scientific Publishing Company.


  • Outlier detection
  • high dimensional data
  • manifold learning

ASJC Scopus subject areas

  • Software
  • Computer Vision and Pattern Recognition
  • Artificial Intelligence


Dive into the research topics of 'Outlier detection in the framework of dimensionality reduction'. Together they form a unique fingerprint.

Cite this