Fitness functions in editing k-NN reference set by genetic algorithms

被引：58

作者：

Kuncheva, LI ^{[1
]}

机构：

[1] UNIV LONDON IMPERIAL COLL SCI TECHNOL & MED,DEPT ELECT & ELECT ENGN,LONDON SW7 2BT,ENGLAND

来源：

PATTERN RECOGNITION | 1997年 / 30卷 / 06期

关键词：

k-Nearest Neighbors (k-NN) rule; genetic algorithms; fitness functions; editing strategies;

D O I：

10.1016/S0031-3203(96)00134-3

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

In a previous paper the use of GAs as an editing technique for the k-nearest neighbor (k-NN) classification technique has been suggested. Here we are looking at different fitness functions. An experimental study with the IRIS data set and with a medical data set has been carried out. Best results (smallest subsets with highest test classification accuracy) have been obtained by including in the fitness function a penalizing term accounting for the cardinality of the reference set. (C) 1997 Pattern Recognition Society.

引用

页码：1041 / 1049

页数：9

共 16 条

[1]

[Anonymous], 1982, Pattern recognition: A statistical approach

[2]

Dasarathy B.V., 1990, NEAREST NEIGHBOR NN

[3] MINIMAL CONSISTENT SET (MCS) IDENTIFICATION FOR OPTIMAL NEAREST-NEIGHBOR DECISION SYSTEMS-DESIGN [J].

DASARATHY, BV .

IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS, 1994, 24 (03) :511-517

[4] A MULTICLASS, K-NN APPROACH TO BAYES RISK-ESTIMATION [J].

DEVIJVER, PA .

PATTERN RECOGNITION LETTERS, 1985, 3 (01) :1-6

[5]

DEVIJVER PA, 1980, 5TH P INT C PATT REC, P72

[6] ADDITIVE ESTIMATORS FOR PROBABILITIES OF CORRECT CLASSIFICATION [J].

GLICK, N .

PATTERN RECOGNITION, 1978, 10 (03) :211-222

[7]

Goldberg DE, 1989, GENETIC ALGORITHMS S

[8]

Kittler J., 1982, Proceedings of the 6th International Conference on Pattern Recognition, P868

[9] EDITING FOR THE K-NEAREST NEIGHBORS RULE BY A GENETIC ALGORITHM [J].

KUNCHEVA, LI .

PATTERN RECOGNITION LETTERS, 1995, 16 (08) :809-814

[10]

KUNCHEVA LL, 1994, P 2 EUR C FUZZ INT T, P640

← 1 2 →