Determining the structure of decision directed acyclic graphs for multiclass classification problems

Determining the structure of decision directed acyclic graphs for multiclass classification problems

Author Quiterio, Thaise M. Autor UNIFESP Google Scholar
Autor UNIFESP Google Scholar
Abstract An usual strategy to solve multiclass classification problems in Machine Learning is to decompose them into multiple binary sub-problems. The final multiclass prediction is obtained by a proper combination of the outputs of the binary classifiers induced in their solution. Decision directed acyclic graphs (DDAG) can be used to organize and to aggregate the outputs of the pairwise classifiers from the one-versus-one (OVO) decomposition. Nonetheless, there are various possible DDAG structures for problems with many classes. In this paper evolutionary algorithms are employed to heuristically find the positions of the OVO binary classifiers in a DDAG. The objective is to place easier sub-problems at higher levels of the DDAG hierarchical structure, in order to minimize the occurrence of cumulative errors. For estimating the complexity of the binary sub-problems, we employ two indexes which measure the separability of the classes. The proposed approach presented sound results in a set of experiments on benchmark datasets, although random DDAGs also performed quite well.
Keywords Recognition
Language English
Sponsor FAPESP [2015/17291-3, 2012/22608-8]
CNPq
Grant number FAPESP: 2015/17291-3
FAPESP: 2012/22608-8
Date 2016
Published in Proceedings Of 2016 5th Brazilian Conference On Intelligent Systems (BRACIS 2016). New york, p. 115-120, 2016.
Publisher Revista De Saude Publica
Extent 115-120
Origin http://dx.doi.org/10.1109/BRACIS.2016.21
Access rights Closed access
Type Conference paper
Web of Science ID WOS:000401813700020
URI http://repositorio.unifesp.br/handle/11600/49243

Show full item record




File

File Size Format View

There are no files associated with this item.

This item appears in the following Collection(s)

Search


Browse

Statistics

My Account