@incollection{fdi:010069089,
  title = {{D}eep {S}elf-{O}rganising {M}aps for efficient heterogeneous biomedical signatures extraction},
  author = {{S}okolovska, {N}. and {N}guyen {T}hanh {H}ai and {C}l{\'e}ment, {K}. and {Z}ucker, {J}ean-{D}aniel},
  editor = {},
  language = {{ENG}},
  abstract = {{F}eature selection is used to preserve significant properties of data in a compact space. {I}n particular, feature selection is needed in applications, where information comes from multiple heterogeneous high dimensional sources. {D}ata integration, however, is a challenge in itself. {I}n 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. {F}or 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. {W}e 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.},
  keywords = {},
  booktitle = {{P}roceedings of the 2016 {I}nternational {J}oint {C}onference on {N}eural {N}etworks ({IJCNN})},
  numero = {},
  pages = {5079--5086},
  address = {{P}iscataway},
  publisher = {{IEEE}},
  series = {},
  year = {2016},
  DOI = {10.1109/{IJCNN}.2016.7727869},
  ISBN = {978-1-5090-0619-9},
  URL = {https://www.documentation.ird.fr/hor/fdi:010069089},
}