Multi-class Boosting for imbalanced data.

Fernández Baldera, Antonio and Buenaposada Biencinto, José Miguel and Baumela Molina, Luis (2015). Multi-class Boosting for imbalanced data.. In: "Iberian Conference on Pattern Recognition and Image Analysis", 17-19 Jun 2015, Santiago de Compostela. ISBN 978-3-319-19390-8. pp. 57-64. https://doi.org/10.1007/978-3-319-19390-8_7.

Description

Title: Multi-class Boosting for imbalanced data.
Author/s:
  • Fernández Baldera, Antonio
  • Buenaposada Biencinto, José Miguel
  • Baumela Molina, Luis
Item Type: Presentation at Congress or Conference (Article)
Event Title: Iberian Conference on Pattern Recognition and Image Analysis
Event Dates: 17-19 Jun 2015
Event Location: Santiago de Compostela
Title of Book: Pattern Recognition and Image Analysis
Date: 2015
ISBN: 978-3-319-19390-8
Volume: 9117
Subjects:
Freetext Keywords: Boosting; Multi-class classification; Imbalanced data
Faculty: E.T.S. de Ingenieros Informáticos (UPM)
Department: Inteligencia Artificial
Creative Commons Licenses: Recognition - No derivative works - Non commercial

Full text

[img]
Preview
PDF - Requires a PDF viewer, such as GSview, Xpdf or Adobe Acrobat Reader
Download (217kB) | Preview

Abstract

We consider the problem of multi-class classification with imbalanced data-sets. To this end, we introduce a cost-sensitive multi-class Boosting algorithm (BAdaCost) based on a generalization of the Boosting margin, termed multi-class cost-sensitive margin. To address the class imbalance we introduce a cost matrix that weighs more hevily the costs of confused classes and a procedure to estimate these costs from the confusion matrix of a standard 0|1-loss classifier. Finally, we evaluate the performance of the approach with synthetic and real data-sets and compare our results with the AdaC2.M1 algorithm.

Funding Projects

TypeCodeAcronymLeaderTitle
Government of SpainTIN2013-47630-C2-2-RUnspecifiedUniversidad Politécnica de MadridSupervisión de patrones de comportamiento humano mediante visión por computador

More information

Item ID: 44043
DC Identifier: http://oa.upm.es/44043/
OAI Identifier: oai:oa.upm.es:44043
DOI: 10.1007/978-3-319-19390-8_7
Official URL: http://www.springer.com/us/book/9783319193892
Deposited by: Memoria Investigacion
Deposited on: 11 Jan 2017 10:33
Last Modified: 27 Nov 2018 09:05
  • Logo InvestigaM (UPM)
  • Logo GEOUP4
  • Logo Open Access
  • Open Access
  • Logo Sherpa/Romeo
    Check whether the anglo-saxon journal in which you have published an article allows you to also publish it under open access.
  • Logo Dulcinea
    Check whether the spanish journal in which you have published an article allows you to also publish it under open access.
  • Logo de Recolecta
  • Logo del Observatorio I+D+i UPM
  • Logo de OpenCourseWare UPM