%0 Conference Proceedings
%T Introducing the Separability Matrix for Error Correcting Output Codes Coding
%A Miguel Angel Bautista
%A Oriol Pujol
%A Xavier Baro
%A Sergio Escalera
%E Carlo Sansone
%E Josef Kittler
%E Fabio Roli
%B 10th International Conference on Multiple Classifier Systems
%D 2011
%V 6713
%I Springer-Verlag Berlin, Heidelberg
%@ 0302-9743
%@ 978-3-642-21556-8
%F Miguel Angel Bautista2011
%O MILAB; OR;HuPBA;MV
%O exported from refbase (http://refbase.cvc.uab.es/show.php?record=1887), last updated on Tue, 11 Mar 2014 15:32:32 +0100
%X Error Correcting Output Codes (ECOC) have demonstrate to be a powerful tool for treating multi-class problems. Nevertheless, predefined ECOC designs may not benefit from Error-correcting principles for particular multi-class data. In this paper, we introduce the Separability matrix as a tool to study and enhance designs for ECOC coding. In addition, a novel problem-dependent coding design based on the Separability matrix is tested over a wide set of challenging multi-class problems, obtaining very satisfactory results.
%U http://dl.acm.org/citation.cfm?id=2040929
%P 227-236