Diversity of classifiers and its applications to combination

Lei Lin; Xiaolong Wang; Jiafeng Liu

Diversity of classifiers and its applications to combination

Lei Lin^*
, Xiaolong Wang
, Jiafeng Liu

^*Corresponding author for this work

Faculty of Computing

Harbin Institute of Technology

Research output: Contribution to journal › Article › peer-review

Abstract

In various application areas of pattern recognition, combining multiple classifiers is regarded as a new method for achieving a substantial gain in performance of systems. This paper discusses the properties of the diversity of classifiers and its applications. A method is presented for combining multiple classifiers based on the diversity. Fusion strategies are discussed for providing a basis for combining classifiers. These combination strategies are experimentally tested on online handwritten Chinese character recognition system and their effectiveness is considered.

Original language	English
Pages (from-to)	33-36
Number of pages	4
Journal	High Technology Letters
Volume	8
Issue number	4
State	Published - Dec 2002

Keywords

Combing multiple classifiers
Diversity
Fusion strategy
Handwritten Chinese character recognition

Cite this

@article{cbc38c33cd6b4d9ea24d2fcedb21a975,

title = "Diversity of classifiers and its applications to combination",

abstract = "In various application areas of pattern recognition, combining multiple classifiers is regarded as a new method for achieving a substantial gain in performance of systems. This paper discusses the properties of the diversity of classifiers and its applications. A method is presented for combining multiple classifiers based on the diversity. Fusion strategies are discussed for providing a basis for combining classifiers. These combination strategies are experimentally tested on online handwritten Chinese character recognition system and their effectiveness is considered.",

keywords = "Combing multiple classifiers, Diversity, Fusion strategy, Handwritten Chinese character recognition",

author = "Lei Lin and Xiaolong Wang and Jiafeng Liu",

year = "2002",

month = dec,

language = "英语",

volume = "8",

pages = "33--36",

journal = "High Technology Letters",

issn = "1006-6748",

publisher = "Institute of Scientific and Technical Information of China ",

number = "4",

}

TY - JOUR

T1 - Diversity of classifiers and its applications to combination

AU - Lin, Lei

AU - Wang, Xiaolong

AU - Liu, Jiafeng

PY - 2002/12

Y1 - 2002/12

N2 - In various application areas of pattern recognition, combining multiple classifiers is regarded as a new method for achieving a substantial gain in performance of systems. This paper discusses the properties of the diversity of classifiers and its applications. A method is presented for combining multiple classifiers based on the diversity. Fusion strategies are discussed for providing a basis for combining classifiers. These combination strategies are experimentally tested on online handwritten Chinese character recognition system and their effectiveness is considered.

AB - In various application areas of pattern recognition, combining multiple classifiers is regarded as a new method for achieving a substantial gain in performance of systems. This paper discusses the properties of the diversity of classifiers and its applications. A method is presented for combining multiple classifiers based on the diversity. Fusion strategies are discussed for providing a basis for combining classifiers. These combination strategies are experimentally tested on online handwritten Chinese character recognition system and their effectiveness is considered.

KW - Combing multiple classifiers

KW - Diversity

KW - Fusion strategy

KW - Handwritten Chinese character recognition

UR - https://www.scopus.com/pages/publications/2442705576

M3 - 文章

AN - SCOPUS:2442705576

SN - 1006-6748

VL - 8

SP - 33

EP - 36

JO - High Technology Letters

JF - High Technology Letters

IS - 4

ER -

Diversity of classifiers and its applications to combination

Abstract

Keywords

Other files and links

Fingerprint

Cite this