Atenção!


O atendimento às questões referentes ao Repositório Institucional será interrompido entre os dias 20 de dezembro de 2024 a 5 de janeiro de 2025.

Pedimos a sua compreensão e aproveitamos para desejar boas festas!

 

Multi-Output Tree Chaining: An Interpretative Modelling and Lightweight Multi-Target Approach

dc.contributor.authorMastelini, Saulo Martiello
dc.contributor.authorda Costa, Victor Guilherme Turrisi
dc.contributor.authorSantana, Everton Jose
dc.contributor.authorNakano, Felipe Kenji
dc.contributor.authorGuido, Rodrigo Capobianco [UNESP]
dc.contributor.authorCerri, Ricardo
dc.contributor.authorBarbon, Sylvio
dc.contributor.institutionUniversidade Estadual de Londrina (UEL)
dc.contributor.institutionUniversidade Federal de São Carlos (UFSCar)
dc.contributor.institutionUniversidade Estadual Paulista (Unesp)
dc.date.accessioned2018-12-11T17:36:57Z
dc.date.available2018-12-11T17:36:57Z
dc.date.issued2018-05-05
dc.description.abstractMulti-target regression (MTR) regards predictive problems with multiple numerical targets. To solve this, machine learning techniques can model solutions treating each target as a separated problem based only on the input features. Nonetheless, modelling inter-target correlation can improve predictive performance. When performing MTR tasks using the statistical dependencies of targets, several approaches put aside the evaluation of each pair-wise correlation between those targets, which may differ for each problem. Besides that, one of the main drawbacks of the current leading MTR method is its high memory cost. In this paper, we propose a novel MTR method called Multi-output Tree Chaining (MOTC) to overcome the mentioned disadvantages. Our method provides an interpretative internal tree-based structure which represents the relationships between targets denominated Chaining Trees (CT). Different from the current techniques, we compute the outputs dependencies, one-by-one, based on the Random Forest importance metric. Furthermore, we proposed a memory friendly approach which reduces the number of required regression models when compared to a leading method, reducing computational cost. We compared the proposed algorithm against three MTR methods (Single-target - ST; Multi-Target Regressor Stacking - MTRS; and Ensemble of Regressor Chains - ERC) on 18 benchmark datasets with two base regression algorithms (Random Forest and Support Vector Regression). The obtained results show that our method is superior to the ST approach regarding predictive performance, whereas, having no significant difference from ERC and MTRS. Moreover, the interpretative tree-based structures built by MOTC pose as great insight on the relationships among targets. Lastly, the proposed solution used significantly less memory than ERC being very similar in predictive performance.en
dc.description.affiliationComputer Science Department State University of Londrina. Rodovia Celso Garcia Cid Km 380 s/n - Campus Universitário
dc.description.affiliationElectrical Engineering Department State University of Londrina. Rodovia Celso Garcia Cid Km 380 s/n - Campus Universitário
dc.description.affiliationDepartment of Computer Science Federal University of São Carlos Rodovia Washington Luís, km 235
dc.description.affiliationInstituto de Biociências Letras e Ciências Exatas Unesp - Univ Estadual Paulista (São Paulo State University), Rua Cristóvão Colombo 2265, Jd Nazareth
dc.description.affiliationUnespInstituto de Biociências Letras e Ciências Exatas Unesp - Univ Estadual Paulista (São Paulo State University), Rua Cristóvão Colombo 2265, Jd Nazareth
dc.format.extent1-25
dc.identifierhttp://dx.doi.org/10.1007/s11265-018-1376-5
dc.identifier.citationJournal of Signal Processing Systems, p. 1-25.
dc.identifier.doi10.1007/s11265-018-1376-5
dc.identifier.file2-s2.0-85046496214.pdf
dc.identifier.issn1939-8115
dc.identifier.issn1939-8018
dc.identifier.lattes6542086226808067
dc.identifier.orcid0000-0002-0924-8024
dc.identifier.scopus2-s2.0-85046496214
dc.identifier.urihttp://hdl.handle.net/11449/179835
dc.language.isoeng
dc.relation.ispartofJournal of Signal Processing Systems
dc.relation.ispartofsjr0,216
dc.rights.accessRightsAcesso aberto
dc.sourceScopus
dc.subjectInterpretative tree structure
dc.subjectMachine learning
dc.subjectMemory-friendly algorithm
dc.subjectMulti-output
dc.subjectMulti-target regression
dc.titleMulti-Output Tree Chaining: An Interpretative Modelling and Lightweight Multi-Target Approachen
dc.typeArtigo
unesp.author.lattes6542086226808067[5]
unesp.author.orcid0000-0002-0092-3572[1]
unesp.author.orcid0000-0002-0924-8024[5]
unesp.campusUniversidade Estadual Paulista (Unesp), Instituto de Biociências Letras e Ciências Exatas, São José do Rio Pretopt
unesp.departmentCiências da Computação e Estatística - IBILCEpt

Arquivos

Pacote Original

Agora exibindo 1 - 1 de 1
Carregando...
Imagem de Miniatura
Nome:
2-s2.0-85046496214.pdf
Tamanho:
2.59 MB
Formato:
Adobe Portable Document Format
Descrição: