An efficient design of a nearest neighbor classifier for various-scale problems

Heesung Lee, Sungjun Hong, Imran Fareed Nizami, Euntai Kim

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

By appropriate editing of the reference set and judicious selection of features, we can obtain an optimal nearest neighbor (NN) classifier that maximizes the accuracy of classification and saves computational time and memory resources. In this paper, we propose a new method for simultaneous reference set editing and feature selection for a nearest neighbor classifier. The proposed method is based on the genetic algorithm and employs different genetic encoding strategies according to the size of the problem, such that it can be applied to classification problems of various scales. Compared with the conventional methods, the classifier uses some of the considered references and features, not all of them, but demonstrates better classification performance. To demonstrate the performance of the proposed method, we perform experiments on various databases.

Original languageEnglish
Pages (from-to)1020-1027
Number of pages8
JournalPattern Recognition Letters
Volume31
Issue number9
DOIs
Publication statusPublished - 2010 Jul 1

Fingerprint

Classifiers
Feature extraction
Genetic algorithms
Data storage equipment
Experiments

All Science Journal Classification (ASJC) codes

  • Software
  • Signal Processing
  • Computer Vision and Pattern Recognition
  • Artificial Intelligence

Cite this

Lee, Heesung ; Hong, Sungjun ; Nizami, Imran Fareed ; Kim, Euntai. / An efficient design of a nearest neighbor classifier for various-scale problems. In: Pattern Recognition Letters. 2010 ; Vol. 31, No. 9. pp. 1020-1027.
@article{7d112cd60a914b1eba61a9067f6f3fb2,
title = "An efficient design of a nearest neighbor classifier for various-scale problems",
abstract = "By appropriate editing of the reference set and judicious selection of features, we can obtain an optimal nearest neighbor (NN) classifier that maximizes the accuracy of classification and saves computational time and memory resources. In this paper, we propose a new method for simultaneous reference set editing and feature selection for a nearest neighbor classifier. The proposed method is based on the genetic algorithm and employs different genetic encoding strategies according to the size of the problem, such that it can be applied to classification problems of various scales. Compared with the conventional methods, the classifier uses some of the considered references and features, not all of them, but demonstrates better classification performance. To demonstrate the performance of the proposed method, we perform experiments on various databases.",
author = "Heesung Lee and Sungjun Hong and Nizami, {Imran Fareed} and Euntai Kim",
year = "2010",
month = "7",
day = "1",
doi = "10.1016/j.patrec.2010.01.001",
language = "English",
volume = "31",
pages = "1020--1027",
journal = "Pattern Recognition Letters",
issn = "0167-8655",
publisher = "Elsevier",
number = "9",

}

An efficient design of a nearest neighbor classifier for various-scale problems. / Lee, Heesung; Hong, Sungjun; Nizami, Imran Fareed; Kim, Euntai.

In: Pattern Recognition Letters, Vol. 31, No. 9, 01.07.2010, p. 1020-1027.

Research output: Contribution to journalArticle

TY - JOUR

T1 - An efficient design of a nearest neighbor classifier for various-scale problems

AU - Lee, Heesung

AU - Hong, Sungjun

AU - Nizami, Imran Fareed

AU - Kim, Euntai

PY - 2010/7/1

Y1 - 2010/7/1

N2 - By appropriate editing of the reference set and judicious selection of features, we can obtain an optimal nearest neighbor (NN) classifier that maximizes the accuracy of classification and saves computational time and memory resources. In this paper, we propose a new method for simultaneous reference set editing and feature selection for a nearest neighbor classifier. The proposed method is based on the genetic algorithm and employs different genetic encoding strategies according to the size of the problem, such that it can be applied to classification problems of various scales. Compared with the conventional methods, the classifier uses some of the considered references and features, not all of them, but demonstrates better classification performance. To demonstrate the performance of the proposed method, we perform experiments on various databases.

AB - By appropriate editing of the reference set and judicious selection of features, we can obtain an optimal nearest neighbor (NN) classifier that maximizes the accuracy of classification and saves computational time and memory resources. In this paper, we propose a new method for simultaneous reference set editing and feature selection for a nearest neighbor classifier. The proposed method is based on the genetic algorithm and employs different genetic encoding strategies according to the size of the problem, such that it can be applied to classification problems of various scales. Compared with the conventional methods, the classifier uses some of the considered references and features, not all of them, but demonstrates better classification performance. To demonstrate the performance of the proposed method, we perform experiments on various databases.

UR - http://www.scopus.com/inward/record.url?scp=77951111106&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77951111106&partnerID=8YFLogxK

U2 - 10.1016/j.patrec.2010.01.001

DO - 10.1016/j.patrec.2010.01.001

M3 - Article

VL - 31

SP - 1020

EP - 1027

JO - Pattern Recognition Letters

JF - Pattern Recognition Letters

SN - 0167-8655

IS - 9

ER -