A Subgraph-Based Selective Classifier Ensemble Algorithm

Abstract:

To improve the generalization performance for ensemble learning, a subgraph based selective classifier ensemble algorithm is presented. Firstly, a set of classifiers are generated by bootstrap sampling technique and support vector machine learning algorithm. And a complete undirected graph is constructed whose vertex is classifier and weight of edge between a pair of classifiers is diversity values. Secondly, by searching technique to find an edge with minimum weight and to calculate similarity values about two vertexes which is related to the edge, vertex with smaller similarity value is removed. According to this method, a subgraph is obtained. Finally, we choose vertexes of subgraph, i.e. classifiers, as ensemble members. Experiments show that presented method outperforms the traditional ensemble learning methods in classification accuracy.

• Periodical: Advanced Materials Research (Volumes 219-220)
• Pages: 261-264
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.219-220.261
• Online since: March 2011
• Authors: Kai Li, Hong Tao Gao
• Keywords: Diversity, Ensemble Learning, Subgraph, Support Vector Machine (SVM)

References

[1] G Brown: Ensemble Learning. Encyclopedia of Machine Learning. Springer press, Heidelberg (2010).

[2] ZH Zhou, J Wu and W Tang: Ensembling neural networks: Many could be better than all. Artificial Intelligence Vol.137(1/2) (2002), pp.239-263.

DOI: 10.1016/s0004-3702(02)00190-x

[3] GZ Li, J Yang and AS Kong: Clustering Algorithm Based Selective Ensemble. Journal of Fudan University (Natural Science) Vol.43(5) (2004), pp.689-691.

[4] K Li, HT Gao: A Novel Measure of Diversity for Support Vector Machine Ensemble. Third International Symposium on Intelligent Information Technology and Security Informatics (2010), pp.366-370.

DOI: 10.1109/iitsi.2010.136