NONPARAMETRIC ESTIMATION OF THE BAYES ERROR OF FEATURE EXTRACTORS USING ORDERED NEAREST NEIGHBOR SETS.

James M. Garnett, Sik-Sang Yau

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

Since the Bayes classifier is the optimum classifier in the sense of having minimum probability of misclassification among all the classifiers using the same set of pattern features, the error rate of the Bayes classifier using the set of features provided by a feature extractor, called the Bayes error of the feature extractor, is the smallest possible for the feature extractor. Consequently, the Bayes error can be used to evaluate the effectiveness of the feature extractors in a pattern recognition system. In this paper, a nonparametric technique for estimating the Bayes error for any two-category feature extractor is presented. This technique uses the nearest neighbor sample sets and is based on an infinite series expansion of the general form of the Bayes error. It is shown that this technique is better than the existing methods, and the estimates obtained by this technique are more meaningful in evaluating the quality of feature extractors. Computer simulation as well as application to electrocardiogram analysis are used to demonstrate this technique.

Original languageEnglish (US)
Pages (from-to)46-54
Number of pages9
JournalIEEE Transactions on Computers
VolumeC-26
Issue number1
StatePublished - Jan 1977
Externally publishedYes

Fingerprint

Extractor
Nonparametric Estimation
Bayes
Nearest Neighbor
Classifiers
Bayes Classifier
Pattern recognition systems
Classifier
Probability of Misclassification
Computer Applications
Electrocardiography
Infinite series
Series Expansion
Pattern Recognition
Error Rate
Computer Simulation
Computer simulation
Evaluate
Estimate
Demonstrate

ASJC Scopus subject areas

  • Hardware and Architecture
  • Electrical and Electronic Engineering

Cite this

NONPARAMETRIC ESTIMATION OF THE BAYES ERROR OF FEATURE EXTRACTORS USING ORDERED NEAREST NEIGHBOR SETS. / Garnett, James M.; Yau, Sik-Sang.

In: IEEE Transactions on Computers, Vol. C-26, No. 1, 01.1977, p. 46-54.

Research output: Contribution to journalArticle

@article{88d1243473eb4931a5c4a457f34f6102,
title = "NONPARAMETRIC ESTIMATION OF THE BAYES ERROR OF FEATURE EXTRACTORS USING ORDERED NEAREST NEIGHBOR SETS.",
abstract = "Since the Bayes classifier is the optimum classifier in the sense of having minimum probability of misclassification among all the classifiers using the same set of pattern features, the error rate of the Bayes classifier using the set of features provided by a feature extractor, called the Bayes error of the feature extractor, is the smallest possible for the feature extractor. Consequently, the Bayes error can be used to evaluate the effectiveness of the feature extractors in a pattern recognition system. In this paper, a nonparametric technique for estimating the Bayes error for any two-category feature extractor is presented. This technique uses the nearest neighbor sample sets and is based on an infinite series expansion of the general form of the Bayes error. It is shown that this technique is better than the existing methods, and the estimates obtained by this technique are more meaningful in evaluating the quality of feature extractors. Computer simulation as well as application to electrocardiogram analysis are used to demonstrate this technique.",
author = "Garnett, {James M.} and Sik-Sang Yau",
year = "1977",
month = "1",
language = "English (US)",
volume = "C-26",
pages = "46--54",
journal = "IEEE Transactions on Computers",
issn = "0018-9340",
publisher = "IEEE Computer Society",
number = "1",

}

TY - JOUR

T1 - NONPARAMETRIC ESTIMATION OF THE BAYES ERROR OF FEATURE EXTRACTORS USING ORDERED NEAREST NEIGHBOR SETS.

AU - Garnett, James M.

AU - Yau, Sik-Sang

PY - 1977/1

Y1 - 1977/1

N2 - Since the Bayes classifier is the optimum classifier in the sense of having minimum probability of misclassification among all the classifiers using the same set of pattern features, the error rate of the Bayes classifier using the set of features provided by a feature extractor, called the Bayes error of the feature extractor, is the smallest possible for the feature extractor. Consequently, the Bayes error can be used to evaluate the effectiveness of the feature extractors in a pattern recognition system. In this paper, a nonparametric technique for estimating the Bayes error for any two-category feature extractor is presented. This technique uses the nearest neighbor sample sets and is based on an infinite series expansion of the general form of the Bayes error. It is shown that this technique is better than the existing methods, and the estimates obtained by this technique are more meaningful in evaluating the quality of feature extractors. Computer simulation as well as application to electrocardiogram analysis are used to demonstrate this technique.

AB - Since the Bayes classifier is the optimum classifier in the sense of having minimum probability of misclassification among all the classifiers using the same set of pattern features, the error rate of the Bayes classifier using the set of features provided by a feature extractor, called the Bayes error of the feature extractor, is the smallest possible for the feature extractor. Consequently, the Bayes error can be used to evaluate the effectiveness of the feature extractors in a pattern recognition system. In this paper, a nonparametric technique for estimating the Bayes error for any two-category feature extractor is presented. This technique uses the nearest neighbor sample sets and is based on an infinite series expansion of the general form of the Bayes error. It is shown that this technique is better than the existing methods, and the estimates obtained by this technique are more meaningful in evaluating the quality of feature extractors. Computer simulation as well as application to electrocardiogram analysis are used to demonstrate this technique.

UR - http://www.scopus.com/inward/record.url?scp=0017449627&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0017449627&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0017449627

VL - C-26

SP - 46

EP - 54

JO - IEEE Transactions on Computers

JF - IEEE Transactions on Computers

SN - 0018-9340

IS - 1

ER -