Coelho, Frederico
[Universidade Federal de Minas Gerais, Brazil]
Braga, Antonio Padua
[Universidade Federal de Minas Gerais, Brazil]
Verleysen, Michel
[UCL]
In this work the principle of homogeneity between labels and data clusters is exploited in order to develop a semi-supervised Feature Selection method. This principle permits the use of cluster information to improve the estimation of feature relevance in order to increase selection performance. Mutual Information is used in a Forward-Backward search process in order to evaluate the relevance of each feature to the data distribution and the existent labels, in a context of few labeled and many unlabeled instances.
| Bibliographic reference |
Coelho, Frederico ; Braga, Antonio Padua ; Verleysen, Michel. Cluster homogeneity as a semi-supervised principle for feature selection using mutual information.20th International Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2012) (Bruges (Belgium), du 25/04/2012 au 27/04/2012). In: Proceedings of the 20th International Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning (ESANN 2012), i6doc.com2012, p. 507-512 |
| Permanent URL |
http://hdl.handle.net/2078.1/116654 |
