Skip to main navigation Skip to search Skip to main content

KPCA-based node selection for fast KMSE

  • Hong Kong Polytechnic University

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this paper we first show that the kernel minimum squared error model is not computationally efficient in feature extraction. To speed up the feature extraction, we linearly express the feature extractor using nodes, i.e. a portion of the training samples in the kernel space. For node selection from the training set, we define two criteria based on Kernel principal component analysis. The nodes are representative and not similar to each other. The experimental results show the feasibility of the proposed method.

Original languageEnglish
Title of host publicationProceedings of the 2014 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2014
EditorsHamid R. Arabnia, Leonidas Deligiannidis, Joan Lu, Fernando G. Tinetti, Jane You, George Jandieri, Gerald Schaefer, Ashu M. G. Solo
PublisherCSREA Press
Pages165-169
Number of pages5
ISBN (Electronic)1601322801, 9781601322807
StatePublished - 2014
Externally publishedYes
Event2014 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2014, at WORLDCOMP 2014 - Las Vegas, United States
Duration: 21 Jul 201424 Jul 2014

Publication series

NameProceedings of the 2014 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2014

Conference

Conference2014 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2014, at WORLDCOMP 2014
Country/TerritoryUnited States
CityLas Vegas
Period21/07/1424/07/14

Keywords

  • Feature Extraction
  • Kernel Minimum Squared Error
  • Minimum Squared Error
  • Pattern Recognition

Fingerprint

Dive into the research topics of 'KPCA-based node selection for fast KMSE'. Together they form a unique fingerprint.

Cite this