Efficient Iteratively reweighted LASSO algorithm for cross-products penalized sparse solutions

Ver/Abrir
Identificadores
URI: http://hdl.handle.net/10902/21904
ISBN: 978-9-0827-9705-3
Compartir
Estadísticas
Ver Estadísticas
Google Scholar
Registro completo
Mostrar el registro completo DC
Autoría
Luengo García, David; Vía Rodríguez, JavierAutoridad Unican; Trigano, Thomas
Fecha
2020
Derechos
© EURASIP. First published in the Proceedings of the 28th European Signal Processing Conference (EUSIPCO-2020) in 2020, published by EURASIP. IEEE is granted the nonexclusive, irrevocable, royalty-free worldwide rights to publish, sell and distribute the copyrighted work in any format or media without restriction.
Publicado en
28th European Signal Processing Conference (EUSIPCO), Amsterdam, Netherlands, 2020, 2045-2049
Editorial
Institute of Electrical and Electronics Engineers, Inc.
Enlace a la publicación
Palabras clave
Sparsity-aware learning
LASSO
Sparse coding
Non-convex optimization
Resumen/Abstract
In this paper, we describe an efficient iterative algorithm for finding sparse solutions to a linear system. Apart from the well-known L1 norm regularization, we introduce an additional cost term promoting solutions without too-close activations. This additional term, which is expressed as a sum of cross-products of absolute values, makes the problem nonconvex and difficult to solve. However, the application of the successive convex approximations approach allows us to obtain an efficient algorithm consisting in the solution of a sequence of iteratively reweighted LASSO problems. Numerical simulations on randomly generated waveforms and ECG signals show the good performance of the proposed method.
Colecciones a las que pertenece