The class imbalance problem in UCS classifier system: A preliminary study

Albert Orriols-Puig; Ester Bernadó-Mansilla

doi:10.1007/978-3-540-71231-2_12

The class imbalance problem in UCS classifier system: A preliminary study

Albert Orriols-Puig^*
, Ester Bernadó-Mansilla

^*Autor/a de correspondencia de este trabajo

Producción científica: Capítulo del libro › Contribución a congreso/conferencia › revisión exhaustiva

7 Citas (Scopus)

Resumen

The class imbalance problem has been said recently to hinder the performance of learning systems. In fact, many of them are designed with the assumption of well-balanced datasets. But this commitment is not always true, since it is very common to find higher presence of one of the classes in real classification problems. The aim of this paper is to make a preliminary analysis on the effect of the class imbalance problem in learning classifier systems. Particularly we focus our study on UCS, a supervised version of XCS classifier system. We analyze UCS's behavior on unbalanced dataseis and find that UCS is sensitive to high levels of class imbalance. We study strategies for dealing with class imbalances, acting either at the sampling level or at the classifier system's level.

Idioma original	Inglés
Título de la publicación alojada	Learning Classifier Systems - International Workshops, IWLCS 2003-2005, Revised Selected Papers
Editorial	Springer Verlag
Páginas	161-180
Número de páginas	20
ISBN (versión impresa)	9783540712305
DOI	https://doi.org/10.1007/978-3-540-71231-2_12
Estado	Publicada - 2007

Serie de la publicación

Nombre	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volumen	4399 LNAI
ISSN (versión impresa)	0302-9743
ISSN (versión digital)	1611-3349

Acceder al documento

10.1007/978-3-540-71231-2_12

Otros archivos y enlaces

Enlace a la publicación en Scopus

Cómo citar

Orriols-Puig, A., & Bernadó-Mansilla, E. (2007). The class imbalance problem in UCS classifier system: A preliminary study. En Learning Classifier Systems - International Workshops, IWLCS 2003-2005, Revised Selected Papers (pp. 161-180). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4399 LNAI). Springer Verlag. https://doi.org/10.1007/978-3-540-71231-2_12

Orriols-Puig, Albert ; Bernadó-Mansilla, Ester. / The class imbalance problem in UCS classifier system : A preliminary study. Learning Classifier Systems - International Workshops, IWLCS 2003-2005, Revised Selected Papers. Springer Verlag, 2007. pp. 161-180 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{3ce36bb95907403a8a88053f36489f0a,

title = "The class imbalance problem in UCS classifier system: A preliminary study",

abstract = "The class imbalance problem has been said recently to hinder the performance of learning systems. In fact, many of them are designed with the assumption of well-balanced datasets. But this commitment is not always true, since it is very common to find higher presence of one of the classes in real classification problems. The aim of this paper is to make a preliminary analysis on the effect of the class imbalance problem in learning classifier systems. Particularly we focus our study on UCS, a supervised version of XCS classifier system. We analyze UCS's behavior on unbalanced dataseis and find that UCS is sensitive to high levels of class imbalance. We study strategies for dealing with class imbalances, acting either at the sampling level or at the classifier system's level.",

author = "Albert Orriols-Puig and Ester Bernad{\'o}-Mansilla",

year = "2007",

doi = "10.1007/978-3-540-71231-2\_12",

language = "English",

isbn = "9783540712305",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "161--180",

booktitle = "Learning Classifier Systems - International Workshops, IWLCS 2003-2005, Revised Selected Papers",

address = "Germany",

}

Orriols-Puig, A & Bernadó-Mansilla, E 2007, The class imbalance problem in UCS classifier system: A preliminary study. En Learning Classifier Systems - International Workshops, IWLCS 2003-2005, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4399 LNAI, Springer Verlag, pp. 161-180. https://doi.org/10.1007/978-3-540-71231-2_12

The class imbalance problem in UCS classifier system: A preliminary study. / Orriols-Puig, Albert; Bernadó-Mansilla, Ester.
Learning Classifier Systems - International Workshops, IWLCS 2003-2005, Revised Selected Papers. Springer Verlag, 2007. p. 161-180 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4399 LNAI).

Producción científica: Capítulo del libro › Contribución a congreso/conferencia › revisión exhaustiva

TY - GEN

T1 - The class imbalance problem in UCS classifier system

T2 - A preliminary study

AU - Orriols-Puig, Albert

AU - Bernadó-Mansilla, Ester

PY - 2007

Y1 - 2007

N2 - The class imbalance problem has been said recently to hinder the performance of learning systems. In fact, many of them are designed with the assumption of well-balanced datasets. But this commitment is not always true, since it is very common to find higher presence of one of the classes in real classification problems. The aim of this paper is to make a preliminary analysis on the effect of the class imbalance problem in learning classifier systems. Particularly we focus our study on UCS, a supervised version of XCS classifier system. We analyze UCS's behavior on unbalanced dataseis and find that UCS is sensitive to high levels of class imbalance. We study strategies for dealing with class imbalances, acting either at the sampling level or at the classifier system's level.

AB - The class imbalance problem has been said recently to hinder the performance of learning systems. In fact, many of them are designed with the assumption of well-balanced datasets. But this commitment is not always true, since it is very common to find higher presence of one of the classes in real classification problems. The aim of this paper is to make a preliminary analysis on the effect of the class imbalance problem in learning classifier systems. Particularly we focus our study on UCS, a supervised version of XCS classifier system. We analyze UCS's behavior on unbalanced dataseis and find that UCS is sensitive to high levels of class imbalance. We study strategies for dealing with class imbalances, acting either at the sampling level or at the classifier system's level.

UR - https://www.scopus.com/pages/publications/38049103501

U2 - 10.1007/978-3-540-71231-2_12

DO - 10.1007/978-3-540-71231-2_12

M3 - Conference contribution

AN - SCOPUS:38049103501

SN - 9783540712305

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

SP - 161

EP - 180

BT - Learning Classifier Systems - International Workshops, IWLCS 2003-2005, Revised Selected Papers

PB - Springer Verlag

ER -

Orriols-Puig A, Bernadó-Mansilla E. The class imbalance problem in UCS classifier system: A preliminary study. En Learning Classifier Systems - International Workshops, IWLCS 2003-2005, Revised Selected Papers. Springer Verlag. 2007. p. 161-180. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-540-71231-2_12

The class imbalance problem in UCS classifier system: A preliminary study

Resumen

Serie de la publicación

Acceder al documento

Otros archivos y enlaces

Huella

Cómo citar