Using sliced mean variance-covariance inverse regression for classification and dimension reduction

Charles D. Lindsey, Simon J. Sheather, Joseph W. McKean

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


The sliced mean variance-covariance inverse regression (SMVCIR) algorithm takes grouped multivariate data as input and transforms it to a new coordinate system where the group mean, variance, and covariance differences are more apparent. Other popular algorithms used for performing graphical group discrimination are sliced average variance estimation (SAVE, targetting the same differences but using a different arrangement for variances) and sliced inverse regression (SIR, which targets mean differences). We provide an improved SMVCIR algorithm and create a dimensionality test for the SMVCIR coordinate system. Simulations corroborating our theoretical results and comparing SMVCIR with the other methods are presented. We also provide examples demonstrating the use of SMVCIR and the other methods, in visualization and group discrimination by k-nearest neighbors. The advantages and differences of SMVCIR from SAVE and SIR are shown clearly in these examples and simulation.

Original languageEnglish
Pages (from-to)769-798
Number of pages30
JournalComputational Statistics
Issue number3-4
StatePublished - Jun 2014


  • Discrimination
  • SAVE
  • SIR
  • Singular value decomposition
  • Visualization

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty
  • Computational Mathematics


Dive into the research topics of 'Using sliced mean variance-covariance inverse regression for classification and dimension reduction'. Together they form a unique fingerprint.

Cite this