An almost surely optimal combined classification rule

被引：6

作者：

Mojirsheibani, M ^{[1
]}

机构：

[1] Carleton Univ, Ottawa, ON K1S 5B6, Canada

来源：

JOURNAL OF MULTIVARIATE ANALYSIS | 2002年 / 81卷 / 01期

基金：

加拿大自然科学与工程研究理事会;

关键词：

Bayes rule; misclassification error; consistencey; Vapnik-Chervonenkis;

D O I：

10.1006/jmva.2001.1990

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 ; 070103 ; 0714 ;

摘要：

We propose a data-based procedure for combining a number of individual classifiers in order to construct more effective classification rules. Under some regularity conditions, the resulting combined classifier turns out to be almost surely superior to each of the individual classifiers. Here, superiority means lower misclassification error rate. (C) 2002 Elsevier Science (USA).

引用

页码：28 / 46

页数：19

共 12 条

[1] MULTIVARIATE BINARY DISCRIMINATION BY KERNEL METHOD [J].

AITCHISON, J ;

AITKEN, CGG .

BIOMETRIKA, 1976, 63 (03) :413-420

[2]

[Anonymous], 1996, PATTERN CLASSIFICATI

[3]

Breiman L, 1996, MACH LEARN, V24, P49

[4]

Devroye L., 1996, A probabilistic theory of pattern recognition

[5]

HALL P, 1981, BIOMETRIKA, V68, P287, DOI 10.1093/biomet/68.1.287

[6]

HILLS M, 1967, J ROYAL STAT SOC B, V28, P1

[7] Combining estimates in regression and classification [J].

LeBlanc, M ;

Tibshirani, R .

JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1996, 91 (436) :1641-1650

[8]

Lugosi G, 1996, ANN STAT, V24, P687

[9]

Mojirsheibani M, 1999, J AM STAT ASSOC, V94, P600

[10] A consistent combined classification rule [J].

Mojirsheibani, M .

STATISTICS & PROBABILITY LETTERS, 1997, 36 (01) :43-47

← 1 2 →