Asymptotic properties of the misclassification rates for Euclidean distance discriminant rule in high-dimensional data

DOI10.1016/J.JMVA.2015.05.008zbMATH Open1327.62383arXiv1403.0329OpenAlexW2964269195MaRDI QIDQ495367FDOQ495367

Masashi Hyodo, Hiroki Watanabe, Tatjana Pavlenko, Takashi Seo

Publication date: 10 September 2015

Published in: Journal of Multivariate Analysis (Search for Journal in Brave)

Abstract: Performance accuracy of the Euclidean Distance Discriminant rule (EDDR) is studied in the high-dimensional asymptotic framework which allows the dimensionality to exceed sample size. Under mild assumptions on the traces of the covariance matrix, our new results provide the asymptotic distribution of the conditional misclassification error and the explicit expression for the consistent and asymptotically unbiased estimator of the expected misclassification error. To get these properties, new results on the asymptotic normality of the quadratic forms and traces of the higher power of Wishart matrix, are established. Using our asymptotic results, we further develop two generic methods of determining a cut-off point for EDDR to adjust the misclassification errors. Finally, we numerically justify the high accuracy of our asymptotic findings along with the cut-off determination methods in finite sample applications, inclusive of the large sample and high-dimensional scenarios.

Full work available at URL: https://arxiv.org/abs/1403.0329

zbMATH Keywords

high-dimensional frameworkexpected error rateconditional error rate

Mathematics Subject Classification ID

Multivariate distribution of statistics (62H10) Classification and discrimination; cluster analysis (statistical aspects) (62H30) Asymptotic distribution theory in statistics (62E20) Estimation in multivariate analysis (62H12)

Cites Work

• An asymptotic expansion of the distribution of the studentized classification statistic W
• A distance-based, misclassification rate adjusted classifier for multiclass, high-dimensional data
• Constrained linear discriminant rule via the Studentized classification statistic based on monotone missing data
• Constrained sample discrimination with the studentized classification statistic w
• Decision Rules, Based on the Distance, for Problems of Fit, Two Samples, and Estimation
• On the fundamental theorem for the decision rule based on distance || ||
• Minimum distance classification rules for high dimensional data

Cited In (7)

• Distance-based classifier by data transformation for high-dimension, strongly spiked eigenvalue models
• Kick-one-out-based variable selection method for Euclidean distance-based classifier in high-dimensional settings
• Asymptotic approximation of misclassification probabilities in linear discriminant analysis with repeated measurements
• Geometric classifiers for high-dimensional noisy data
• Asymptotic expansions of the probabilities of misclassification for k-NN discriminant rules
• Approximation of misclassification probabilities in linear discriminant analysis based on repeated measurements
• Asymptotic estimate of probability of misclassification for discriminant rules based on density estimates

This page was built for publication: Asymptotic properties of the misclassification rates for Euclidean distance discriminant rule in high-dimensional data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q495367)