Increasing Diversity in Random Forests Using Naive Bayes - Archive ouverte HAL Access content directly
Conference Papers Year : 2016

Increasing Diversity in Random Forests Using Naive Bayes

(1) , (1) , (1)
1
Christos K. Aridas
  • Function : Author
  • PersonId : 1011977
Sotiris B. Kotsiantis
  • Function : Author
  • PersonId : 1011978
Michael N. Vrahatis
  • Function : Author
  • PersonId : 1011979

Abstract

In this work a novel ensemble technique for generating random decision forests is presented. The proposed technique incorporates a Naive Bayes classification model to increase the diversity of the trees in the forest in order to improve the performance in terms of classification accuracy. Experimental results on several benchmark data sets show that the proposed method archives outstanding predictive performance compared to other state-of-the-art ensemble methods.
Fichier principal
Vignette du fichier
430537_1_En_7_Chapter.pdf (226.8 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01557627 , version 1 (06-07-2017)

Licence

Attribution - CC BY 4.0

Identifiers

Cite

Christos K. Aridas, Sotiris B. Kotsiantis, Michael N. Vrahatis. Increasing Diversity in Random Forests Using Naive Bayes. 12th IFIP International Conference on Artificial Intelligence Applications and Innovations (AIAI), Sep 2016, Thessaloniki, Greece. pp.75-86, ⟨10.1007/978-3-319-44944-9_7⟩. ⟨hal-01557627⟩
63 View
481 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More