Benchmarking a reduced multivariate polynomial pattern classifier

Kar Ann Toh, Quoc Long Tran, Dipti Srinivasan

Research output: Contribution to journalArticle

111 Citations (Scopus)

Abstract

A novel method using a reduced multivariate polynomial model has been developed for biometric decision fusion where simplicity and ease of use could be a concern. However, much to our surprise, the reduced model was found to have good classification accuracy for several commonly used data sets from the Web. In this paper, we extend the single output model to a multiple outputs model to handle multiple class problems. The method is particularly suitable for problems with small number of features and large number of examples. Basic component of this polynomial model boils down to construction of new pattern features which are sums of the original features and combination of these new and original features using power and product terms. A linear regularized least-squares predictor is then built using these constructed features. The number of constructed feature terms varies linearly with the order of the polynomial, instead of having a power law in the case of full multivariate polynomials. The method is simple as it amounts to only a few lines of Matlab code. We perform extensive experiments on this reduced model using 42 data sets. Our results compared remarkably well with best reported results of several commonly used algorithms from the literature. Both the classification accuracy and efficiency aspects are reported for this reduced model.

Original languageEnglish
Pages (from-to)740-755
Number of pages16
JournalIEEE transactions on pattern analysis and machine intelligence
Volume26
Issue number6
DOIs
Publication statusPublished - 2004 Jun 1

Fingerprint

Multivariate Polynomials
Reduced Model
Benchmarking
Polynomial Model
Classifiers
Classifier
Polynomials
Decision Fusion
Linear Least Squares
Output
Multivariate Models
Term
Biometrics
MATLAB
Predictors
Simplicity
Power Law
Linearly
Vary
Polynomial

All Science Journal Classification (ASJC) codes

  • Software
  • Computer Vision and Pattern Recognition
  • Computational Theory and Mathematics
  • Artificial Intelligence
  • Applied Mathematics

Cite this

@article{2eac0a14a0564a9091cf881e334b7716,
title = "Benchmarking a reduced multivariate polynomial pattern classifier",
abstract = "A novel method using a reduced multivariate polynomial model has been developed for biometric decision fusion where simplicity and ease of use could be a concern. However, much to our surprise, the reduced model was found to have good classification accuracy for several commonly used data sets from the Web. In this paper, we extend the single output model to a multiple outputs model to handle multiple class problems. The method is particularly suitable for problems with small number of features and large number of examples. Basic component of this polynomial model boils down to construction of new pattern features which are sums of the original features and combination of these new and original features using power and product terms. A linear regularized least-squares predictor is then built using these constructed features. The number of constructed feature terms varies linearly with the order of the polynomial, instead of having a power law in the case of full multivariate polynomials. The method is simple as it amounts to only a few lines of Matlab code. We perform extensive experiments on this reduced model using 42 data sets. Our results compared remarkably well with best reported results of several commonly used algorithms from the literature. Both the classification accuracy and efficiency aspects are reported for this reduced model.",
author = "Toh, {Kar Ann} and Tran, {Quoc Long} and Dipti Srinivasan",
year = "2004",
month = "6",
day = "1",
doi = "10.1109/TPAMI.2004.3",
language = "English",
volume = "26",
pages = "740--755",
journal = "IEEE Transactions on Pattern Analysis and Machine Intelligence",
issn = "0162-8828",
publisher = "IEEE Computer Society",
number = "6",

}

Benchmarking a reduced multivariate polynomial pattern classifier. / Toh, Kar Ann; Tran, Quoc Long; Srinivasan, Dipti.

In: IEEE transactions on pattern analysis and machine intelligence, Vol. 26, No. 6, 01.06.2004, p. 740-755.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Benchmarking a reduced multivariate polynomial pattern classifier

AU - Toh, Kar Ann

AU - Tran, Quoc Long

AU - Srinivasan, Dipti

PY - 2004/6/1

Y1 - 2004/6/1

N2 - A novel method using a reduced multivariate polynomial model has been developed for biometric decision fusion where simplicity and ease of use could be a concern. However, much to our surprise, the reduced model was found to have good classification accuracy for several commonly used data sets from the Web. In this paper, we extend the single output model to a multiple outputs model to handle multiple class problems. The method is particularly suitable for problems with small number of features and large number of examples. Basic component of this polynomial model boils down to construction of new pattern features which are sums of the original features and combination of these new and original features using power and product terms. A linear regularized least-squares predictor is then built using these constructed features. The number of constructed feature terms varies linearly with the order of the polynomial, instead of having a power law in the case of full multivariate polynomials. The method is simple as it amounts to only a few lines of Matlab code. We perform extensive experiments on this reduced model using 42 data sets. Our results compared remarkably well with best reported results of several commonly used algorithms from the literature. Both the classification accuracy and efficiency aspects are reported for this reduced model.

AB - A novel method using a reduced multivariate polynomial model has been developed for biometric decision fusion where simplicity and ease of use could be a concern. However, much to our surprise, the reduced model was found to have good classification accuracy for several commonly used data sets from the Web. In this paper, we extend the single output model to a multiple outputs model to handle multiple class problems. The method is particularly suitable for problems with small number of features and large number of examples. Basic component of this polynomial model boils down to construction of new pattern features which are sums of the original features and combination of these new and original features using power and product terms. A linear regularized least-squares predictor is then built using these constructed features. The number of constructed feature terms varies linearly with the order of the polynomial, instead of having a power law in the case of full multivariate polynomials. The method is simple as it amounts to only a few lines of Matlab code. We perform extensive experiments on this reduced model using 42 data sets. Our results compared remarkably well with best reported results of several commonly used algorithms from the literature. Both the classification accuracy and efficiency aspects are reported for this reduced model.

UR - http://www.scopus.com/inward/record.url?scp=3042573614&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=3042573614&partnerID=8YFLogxK

U2 - 10.1109/TPAMI.2004.3

DO - 10.1109/TPAMI.2004.3

M3 - Article

C2 - 18579935

AN - SCOPUS:3042573614

VL - 26

SP - 740

EP - 755

JO - IEEE Transactions on Pattern Analysis and Machine Intelligence

JF - IEEE Transactions on Pattern Analysis and Machine Intelligence

SN - 0162-8828

IS - 6

ER -