Polynomial Representation of Classifiers with Independent Discrete-Valued Features

Research output: Contribution to journalArticle

Abstract

It is shown that for n-valued conditionally independent features a large family of classifiers can be expressed as an (n—list-degree polynomial discriminant function. The usefulness of the polynomial expansion is discussed and demonstrated by considering the first-order Minkowski metric, the Euclidean distance, and Bayes’ classifiers for the ternary-feature case. Finally, some interesting side observations on the classifiers are made with respect to optimality and computational requirements.

Original languageEnglish (US)
Pages (from-to)205-208
Number of pages4
JournalIEEE Transactions on Computers
VolumeC-21
Issue number2
DOIs
StatePublished - Jan 1 1972

Fingerprint

Classifiers
Classifier
Polynomials
Bayes Classifier
Discriminant Function
Polynomial
Euclidean Distance
Polynomial function
Ternary
Optimality
First-order
Metric
Requirements
Family
Observation

Keywords

  • Bayes’
  • classifier Euclidean distance classifier Minkowski metric classifier polynomial discriminant functions

ASJC Scopus subject areas

  • Software
  • Theoretical Computer Science
  • Hardware and Architecture
  • Computational Theory and Mathematics

Cite this

Polynomial Representation of Classifiers with Independent Discrete-Valued Features. / Toussaint, Godfried.

In: IEEE Transactions on Computers, Vol. C-21, No. 2, 01.01.1972, p. 205-208.

Research output: Contribution to journalArticle

@article{0dd55d593f8c45888b51cdebba880ea1,
title = "Polynomial Representation of Classifiers with Independent Discrete-Valued Features",
abstract = "It is shown that for n-valued conditionally independent features a large family of classifiers can be expressed as an (n—list-degree polynomial discriminant function. The usefulness of the polynomial expansion is discussed and demonstrated by considering the first-order Minkowski metric, the Euclidean distance, and Bayes’ classifiers for the ternary-feature case. Finally, some interesting side observations on the classifiers are made with respect to optimality and computational requirements.",
keywords = "Bayes’, classifier Euclidean distance classifier Minkowski metric classifier polynomial discriminant functions",
author = "Godfried Toussaint",
year = "1972",
month = "1",
day = "1",
doi = "10.1109/TC.1972.5008928",
language = "English (US)",
volume = "C-21",
pages = "205--208",
journal = "IEEE Transactions on Computers",
issn = "0018-9340",
publisher = "IEEE Computer Society",
number = "2",

}

TY - JOUR

T1 - Polynomial Representation of Classifiers with Independent Discrete-Valued Features

AU - Toussaint, Godfried

PY - 1972/1/1

Y1 - 1972/1/1

N2 - It is shown that for n-valued conditionally independent features a large family of classifiers can be expressed as an (n—list-degree polynomial discriminant function. The usefulness of the polynomial expansion is discussed and demonstrated by considering the first-order Minkowski metric, the Euclidean distance, and Bayes’ classifiers for the ternary-feature case. Finally, some interesting side observations on the classifiers are made with respect to optimality and computational requirements.

AB - It is shown that for n-valued conditionally independent features a large family of classifiers can be expressed as an (n—list-degree polynomial discriminant function. The usefulness of the polynomial expansion is discussed and demonstrated by considering the first-order Minkowski metric, the Euclidean distance, and Bayes’ classifiers for the ternary-feature case. Finally, some interesting side observations on the classifiers are made with respect to optimality and computational requirements.

KW - Bayes’

KW - classifier Euclidean distance classifier Minkowski metric classifier polynomial discriminant functions

UR - http://www.scopus.com/inward/record.url?scp=33845276095&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33845276095&partnerID=8YFLogxK

U2 - 10.1109/TC.1972.5008928

DO - 10.1109/TC.1972.5008928

M3 - Article

VL - C-21

SP - 205

EP - 208

JO - IEEE Transactions on Computers

JF - IEEE Transactions on Computers

SN - 0018-9340

IS - 2

ER -