1: \begin{abstract}

2: Sparse modeling is a powerful framework for data analysis and

3: processing. Traditionally, encoding in this framework is done

4: by solving an $\ell_1$-regularized linear regression problem, usually

5: called {\it Lasso}. In this work we first combine

6: the sparsity-inducing property of the Lasso model, at the individual

7: feature level, with the block-sparsity property of the

8: {\it group Lasso} model, where sparse groups of features are jointly encoded, obtaining a sparsity pattern hierarchically

9: structured. This results in the {\it hierarchical

10: Lasso}, which shows important practical modeling advantages. We then

11: extend this approach to the collaborative case, where a set of simultaneously coded signals share the

12: same sparsity pattern at the higher (group) level but not necessarily at the lower

13: one. Signals then share the same active groups, or classes, but not necessarily

14: the same active set. This is very well suited for

15: applications such as source separation. An efficient optimization procedure, which guarantees convergence to the global optimum, is developed for these new models. The underlying presentation of the new framework and optimization approach is complemented with experimental examples and preliminary theoretical results.

16: \end{abstract}

17: