Direct minimization of error rates in multivariate classification

DOI10.1007/S001800200089MaRDI QIDQ1855617zbMATH OpenOpenAlexFDO

Authors Michael C. Röhl, Claus Weihs, Winfried Theis

Publication date 6 February 2003

Published in Computational Statistics (Search for Journal in Brave)

Full work available at URL http://hdl.handle.net/2003/4933

zbMATH Keywords

classificationdimension reductiondiscriminant analysiserror ratesimulated annealinguser preferences

Mathematics Subject Classification ID

Computational methods for problems pertaining to statistics (62-08) Classification and discrimination; cluster analysis (statistical aspects) (62H30)

Review text

The authors propose and discusse two versions of a computer intensive method for linear dimension reduction that minimize the classification error directly. The method is based on a powerful modern optimization technique (simulated annealing optimization algorithm) and is a flexible tool for incorporating user preferences. Simulation and a real world example demonstrate the superiority of the proposed methods over the classical discriminant analysis. Special emphasis is given to the case when discriminant analysis collapses.

Recommendations

• Component reduction in linear discriminant analysis
• Dimension reduction for predictive discrimination
• Dimensionality reduction in quadratic discriminant analysis
• Linear dimension reduction in classification: adaptive procedure for optimum results
• New Fast Algorithms for Error Rate-Based Stepwise Variable Selection in Discriminant Analysis

Cites work

• Generalized Simulated Annealing for Function Optimization
• scientific article; zbMATH DE number 47593 (Why is no real title available?)
• scientific article; zbMATH DE number 2118472 (Why is no real title available?)
• scientific article; zbMATH DE number 274379 (Why is no real title available?)
• Projection pursuit discriminant analysis
• Projection pursuit discriminant analysis for two groups

Cited in

(6)

• Linear dimension reduction in classification: adaptive procedure for optimum results
• A modified minimum classification error (MCE) training algorithm for dimensionality reduction
• Discriminative learning for minimum error classification (pattern recognition)
• Component reduction in linear discriminant analysis
• AN EXTENSION OF A MINIMAX APPROACH TO MULTIPLE CLASSIFICATION
• Generation of prediction optimal projection on latent factors by a stochastic search algorithm

This page was built for publication: Direct minimization of error rates in multivariate classification

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1855617)