Multi-label lazy associative classification

Adriano Veloso, Wagner Meira, Marcos Gonçalves, Mohammed Zaki

Research output: Chapter in Book/Report/Conference proceedingConference contribution

37 Citations (Scopus)

Abstract

Most current work on classification has been focused on learning from a set of instances that are associated with a single label (i.e., single-label classification). However, many applications, such as gene functional prediction and text categorization, may allow the instances to be associated with multiple labels simultaneously. Multi-label classification is a generalization of single-label classification, and its generality makes it much more difficult to solve. Despite its importance, research on multi-label classification is still lacking. Common approaches simply learn independent binary classifiers for each label, and do not exploit dependencies among labels. Also, several small disjuncts may appear due to the possibly large number of label combinations, and neglecting these small disjuncts may degrade classification accuracy. In this paper we propose a multi-label lazy associative classifier, which progressively exploits dependencies among labels. Further, since in our lazy strategy the classification model is induced on an instance-based fashion, the proposed approach can provide a better coverage of small disjuncts. Gains of up to 24% are observed when the proposed approach is compared against the state-of-the-art multi-label classifiers.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages605-612
Number of pages8
Volume4702 LNAI
Publication statusPublished - 1 Dec 2007
Externally publishedYes
Event11th European Conference on Principles and Practice of Knowledge Discovery in Databases, PKDD 2007 - Warsaw, Poland
Duration: 17 Sep 200721 Sep 2007

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4702 LNAI
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other11th European Conference on Principles and Practice of Knowledge Discovery in Databases, PKDD 2007
CountryPoland
CityWarsaw
Period17/9/0721/9/07

Fingerprint

Labels
Classifier
Classifiers
Text Categorization
Coverage
Learning
Binary
Gene
Prediction
Research
Genes

ASJC Scopus subject areas

  • Computer Science(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Theoretical Computer Science

Cite this

Veloso, A., Meira, W., Gonçalves, M., & Zaki, M. (2007). Multi-label lazy associative classification. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4702 LNAI, pp. 605-612). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4702 LNAI).

Multi-label lazy associative classification. / Veloso, Adriano; Meira, Wagner; Gonçalves, Marcos; Zaki, Mohammed.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 4702 LNAI 2007. p. 605-612 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4702 LNAI).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Veloso, A, Meira, W, Gonçalves, M & Zaki, M 2007, Multi-label lazy associative classification. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 4702 LNAI, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4702 LNAI, pp. 605-612, 11th European Conference on Principles and Practice of Knowledge Discovery in Databases, PKDD 2007, Warsaw, Poland, 17/9/07.
Veloso A, Meira W, Gonçalves M, Zaki M. Multi-label lazy associative classification. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 4702 LNAI. 2007. p. 605-612. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Veloso, Adriano ; Meira, Wagner ; Gonçalves, Marcos ; Zaki, Mohammed. / Multi-label lazy associative classification. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 4702 LNAI 2007. pp. 605-612 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{78f9c1cb6a3c4dddbd02fc6bea2a75b5,
title = "Multi-label lazy associative classification",
abstract = "Most current work on classification has been focused on learning from a set of instances that are associated with a single label (i.e., single-label classification). However, many applications, such as gene functional prediction and text categorization, may allow the instances to be associated with multiple labels simultaneously. Multi-label classification is a generalization of single-label classification, and its generality makes it much more difficult to solve. Despite its importance, research on multi-label classification is still lacking. Common approaches simply learn independent binary classifiers for each label, and do not exploit dependencies among labels. Also, several small disjuncts may appear due to the possibly large number of label combinations, and neglecting these small disjuncts may degrade classification accuracy. In this paper we propose a multi-label lazy associative classifier, which progressively exploits dependencies among labels. Further, since in our lazy strategy the classification model is induced on an instance-based fashion, the proposed approach can provide a better coverage of small disjuncts. Gains of up to 24{\%} are observed when the proposed approach is compared against the state-of-the-art multi-label classifiers.",
author = "Adriano Veloso and Wagner Meira and Marcos Gon{\cc}alves and Mohammed Zaki",
year = "2007",
month = "12",
day = "1",
language = "English",
isbn = "9783540749752",
volume = "4702 LNAI",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "605--612",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Multi-label lazy associative classification

AU - Veloso, Adriano

AU - Meira, Wagner

AU - Gonçalves, Marcos

AU - Zaki, Mohammed

PY - 2007/12/1

Y1 - 2007/12/1

N2 - Most current work on classification has been focused on learning from a set of instances that are associated with a single label (i.e., single-label classification). However, many applications, such as gene functional prediction and text categorization, may allow the instances to be associated with multiple labels simultaneously. Multi-label classification is a generalization of single-label classification, and its generality makes it much more difficult to solve. Despite its importance, research on multi-label classification is still lacking. Common approaches simply learn independent binary classifiers for each label, and do not exploit dependencies among labels. Also, several small disjuncts may appear due to the possibly large number of label combinations, and neglecting these small disjuncts may degrade classification accuracy. In this paper we propose a multi-label lazy associative classifier, which progressively exploits dependencies among labels. Further, since in our lazy strategy the classification model is induced on an instance-based fashion, the proposed approach can provide a better coverage of small disjuncts. Gains of up to 24% are observed when the proposed approach is compared against the state-of-the-art multi-label classifiers.

AB - Most current work on classification has been focused on learning from a set of instances that are associated with a single label (i.e., single-label classification). However, many applications, such as gene functional prediction and text categorization, may allow the instances to be associated with multiple labels simultaneously. Multi-label classification is a generalization of single-label classification, and its generality makes it much more difficult to solve. Despite its importance, research on multi-label classification is still lacking. Common approaches simply learn independent binary classifiers for each label, and do not exploit dependencies among labels. Also, several small disjuncts may appear due to the possibly large number of label combinations, and neglecting these small disjuncts may degrade classification accuracy. In this paper we propose a multi-label lazy associative classifier, which progressively exploits dependencies among labels. Further, since in our lazy strategy the classification model is induced on an instance-based fashion, the proposed approach can provide a better coverage of small disjuncts. Gains of up to 24% are observed when the proposed approach is compared against the state-of-the-art multi-label classifiers.

UR - http://www.scopus.com/inward/record.url?scp=38049177740&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=38049177740&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9783540749752

VL - 4702 LNAI

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 605

EP - 612

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -