An analysis of diversity measures

Diversity among the base classifiers is deemed to be important when constructing a classifier ensemble. Numerous algorithms have been proposed to construct a good classifier ensemble by seeking both the accuracy of the base classifiers and the diversity among them. However, there is no generally acc...

Full description

Saved in:
Bibliographic Details
Published in:Machine learning Vol. 65; no. 1; pp. 247 - 271
Main Authors: Tang, E. K., Suganthan, P. N., Yao, X.
Format: Journal Article
Language:English
Published: Dordrecht Springer 01.10.2006
Springer Nature B.V
Subjects:
Applied sciences
Artificial intelligence
Computer science; control theory; systems
Exact sciences and technology
Studies
Diversity measures, Margin distribution, Majority vote
Interrater agreement
Experimental design
Classification
Generalized diversity, Measure of difficulty, Entropy measure, Coincident failure diversity
Classifier ensemble
Disagreement measure, Double fault measure
Learning algorithm
Artificial intelligence
KW variance
Variance
Aggregate model
ISSN:0885-6125, 1573-0565
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Diversity among the base classifiers is deemed to be important when constructing a classifier ensemble. Numerous algorithms have been proposed to construct a good classifier ensemble by seeking both the accuracy of the base classifiers and the diversity among them. However, there is no generally accepted definition of diversity, and measuring the diversity explicitly is very difficult. Although researchers have designed several experimental studies to compare different diversity measures, usually confusing results were observed. In this paper, we present a theoretical analysis on six existing diversity measures (namely disagreement measure, double fault measure, KW variance, inter-rater agreement, generalized diversity and measure of difficulty), show underlying relationships between them, and relate them to the concept of margin, which is more explicitly related to the success of ensemble learning algorithms. We illustrate why confusing experimental results were observed and show that the discussed diversity measures are naturally ineffective. Our analysis provides a deeper understanding of the concept of diversity, and hence can help design better ensemble learning algorithms.[PUBLICATION ABSTRACT]
Bibliography:SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 14
ObjectType-Article-2
content type line 23
ISSN:0885-6125
1573-0565
DOI:10.1007/s10994-006-9449-2