@article{872,
  author = {Hamid Parvin, Alizadeh Hosein, Sajad Parvin},
  title = {A Scalable Classifier Ensemble Suitable for Multiclass Classification by Use of Pairwise Classifier Ensembles},
  journal = {Journal of Intelligent Computing},
  year = {2012},
  volume = {3},
  number = {1},
  doi = {},
  url = {http://www.dline.info/jic/fulltext/v3n1/1.pdf},
  abstract = {To reach the best classification there is a way to use many inaccurate or weak classifiers each of them is specialized for a sub-space in the problem space and using their consensus vote as the final classifier. Many methods have been proposed for classifier ensemble in pattern recognition such as Random Forest which uses a host of decision trees as base classifiers. The paper proposes a heuristic classifier ensemble to improve the performance of learning in the classification. It specially deals with multiclass problems which their aim is to learn the boundaries of each class among many classes. Based on the concept of multiclass problems, the classifiers are divided into two different categories: pairwise classifiers and multiclass classifiers. The aim of a pairwise classifier is to separate one class from another one. Because the pairwise classifiers are just trained to discriminate between two classes, the decision boundaries learned by them are simpler and more effective than those learned by the multiclass classifiers. The two proposed methods are similar to Random Forest method in employing many decision trees and neural networks as base classifiers. For evaluating the proposed weighting methods, both cases of decision tree and neural network classifiers are applied in experimental results. The two proposed ensemble methods are tested on a huge Persian dataset of handwritten digits and it has been shown that the proposed ensemble methods outperforms some other state-of-art ensemble methods.},
}