Home > Published Issues > 2020 > Volume 11, No. 4, November 2020 >

Probability-Weighted Voting Ensemble Learning for Classification Model

Artitayapron Rojarath and Wararat Songpan
Department of Computer Science Faculty of Science, Khon Kaen University, Thailand

Abstract—Many research studies have investigated ensemble learning. However, these research studies proposed an approach for improving the ensemble learning. We propose the efficiency method using probability weight as a support to the classifier model called the probability-weighted voting ensemble learning, which computes its own probability computation for each model from the training data. This research has tested the proposed model with 5 UCI data sets in various dimensions and generated four models, the 3PW-Ensemble model, the 4PW-Ensemble model, the 5PW-Ensemble model, and the 6PW-Ensemble model. The experimental results of the study yield the highest accuracy. Considering the comparison of efficiency, the accuracy of the proposed model was higher than those of the based classification models and the other ensemble models.
Index Terms—classification model, ensemble learning, machine learning, model combination, probability weight, weight voting

Cite: Artitayapron Rojarath and Wararat Songpan, "Probability-Weighted Voting Ensemble Learning for Classification Model," Journal of Advances in Information Technology, Vol. 11, No. 4, pp. 217-227, November 2020. doi: 10.12720/jait.11.4.217-227

Copyright © 2020 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.