Publications des scientifiques de l'IRD

Sokolovska N., Nguyen Thanh Hai, Clément K., Zucker Jean-Daniel. (2016). Deep Self-Organising Maps for efficient heterogeneous biomedical signatures extraction. In : Proceedings of the 2016 International Joint Conference on Neural Networks (IJCNN). Piscataway : IEEE, p. 5079-5086. International Joint Conference on Neural Networks (IJCNN), Vancouver (CAN), 2016/07/24-29. ISBN 978-1-5090-0619-9.

Titre du document
Deep Self-Organising Maps for efficient heterogeneous biomedical signatures extraction
Année de publication
2016
Type de document
Partie d'ouvrage
Auteurs
Sokolovska N., Nguyen Thanh Hai, Clément K., Zucker Jean-Daniel
In
Proceedings of the 2016 International Joint Conference on Neural Networks (IJCNN)
Source
Piscataway : IEEE, 2016, p. 5079-5086 ISBN 978-1-5090-0619-9
Colloque
International Joint Conference on Neural Networks (IJCNN), Vancouver (CAN), 2016/07/24-29
Feature selection is used to preserve significant properties of data in a compact space. In particular, feature selection is needed in applications, where information comes from multiple heterogeneous high dimensional sources. Data integration, however, is a challenge in itself. In our contribution, we introduce a feature selection framework based on powerful visualisation capabilities of self-organising maps, where the deep structure can be learned in a supervised or unsupervised manner. For a supervised version of the deep SOM, we propose to carry out inference with a linear SVM. A forward-backward procedure helps to converge to an optimal feature set. We show by experiments on real large-scale biomedical data set that the proposed methods embed data in a new compact meaningful representation, allow to visualise biomedical signatures, and also lead to a reasonable classification accuracy compared to the state-of-the-art methods.
Plan de classement
Sciences fondamentales / Techniques d'analyse et de recherche [020] ; Santé : généralités [050] ; Biotechnologies [084] ; Informatique [122]
Localisation
Fonds IRD [F B010069089]
Identifiant IRD
fdi:010069089
Contact