Boosting simple projections for multi-class dimensionality reduction

Yuan Yuan*, Yanwei Pang

*Corresponding author for this work

Research output: Chapter in Book/Published conference outputConference publication

Abstract

This paper presents a novel method for dimensionality reduction and for multi-class classification tasks. This method iteratively selects a series of simple but effective 1D subspaces, and then combines the corresponding 1D projections by Adaboost. M2. Its major advantages are: 1) it does not impose speci.c assumptions on data distribution; 2) it minimizes Bayes error estimation in low-dimensional space; 3) it simpli.es existing subspace-based methods to eigenvalue decomposition problem; and 4) each of the 1D subspaces (with associated nearest neighbor classi.er) has different emphasis - measured by weighted training error. Experiments on both synthetic and real-world data demonstrate the effectiveness of the proposed method.

Original languageEnglish
Title of host publicationIEEE International Conference on Systems, Man and Cybernetics, 2008. SMC 2008
PublisherIEEE
Pages2231-2235
Number of pages5
DOIs
Publication statusPublished - 1 Dec 2008

Publication series

NameIEEE international conference on systems, man, and cybernetics : conference proceedings
PublisherIEEE
ISSN (Print)1062-922X

Keywords

  • Adaboost.M2
  • Feature extraction
  • Multi-class classification

Fingerprint

Dive into the research topics of 'Boosting simple projections for multi-class dimensionality reduction'. Together they form a unique fingerprint.

Cite this