Copyright notice: The material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.

If you want a paper that is not available here, or if you have any problems accessing these documents, e-mail me.

k-support reguralized Support Vector Machine

The k-support regularized Support Vector Machine (ksup-SVM) is a spasity regularization method that balances the L1 and L2 norms over a linear function in order to prevent overfilling. This approach enables the learning algorithm to select a sparse but correlated subset of discriminative variables. It has two input parameters, the λ > 0 regularization parameter and k ∈ { 1 , . . . , d }, where d is the dimension of the feature space, the parameter that negatively correlates with the sparsity. Furthermore, the k sup-SVM can be seen as an alternative to the Elastic Net regularized SVM, but with a tighter convex relaxation to correlated sparsity. We evaluated our proposed method in a neurmuscular disease classification task using MRI-based markers.

Publications

  • Sparse classification with MRI based markers for neuromuscular disease categorization.
    Gkirtzou Katerina, Deux Jean-François, Bassez Guillaume, Sotiras Aristeidis, Rahmouni Alain, Varacca Thibault,Paragios Nikos and Blaschko B. Matthew
    4th International Workshop on Machine Learning in Medical Imaging (MLMI), 2013.
    [pdf] [Supplementary Material] [bibtex] [doi]
  • Sparsity regularization and graph-based representation in medical imaging
    Gkirtzou Katerina
    PhD Thesis, Ecole Centrale de Paris, 2013.
    [pdf] [bibtex] [presentation]

Code

The code is provided under the GNU GPL license. It does not come with any warranty of any kind.