Enhancing safe screening rules with adaptive thresholding for non-overlapping group sparse norm regularized problems

Hector Chahuara, Paul Rodriguez

Producción científica: Capítulo del libro/informe/acta de congresoContribución a la conferenciarevisión exhaustiva

Resumen

Sparsity is an often desired property in machine learning and signal processing problems. Recently, techniques such as screening rules were proposed to exploit sparsity in order to diminish the computational requirements of large and huge-scale optimization problems. Nevertheless, existing methods provide rough estimations of the solution support discarding only a few entries in the solution, thus limiting the desired computational savings. In this paper, we propose a simple and computationally cheap modification for safe screening rules based on automatic thresholding and the observation that the screening metric has a distribution that, for practical purposes, can be considered unimodal. The proposed method is evaluated for MEG / EEG source imaging and image classification. Computational results indicate that the proposed screening scheme outperforms the safe method costing only minor losses in accuracy and yields approximate speedups of up to 167.59 for MEG / EEG source imaging, and up to 2.12 for image classification.

Idioma originalInglés
Título de la publicación alojada2023 24th International Conference on Digital Signal Processing, DSP 2023
EditorialInstitute of Electrical and Electronics Engineers Inc.
ISBN (versión digital)9798350339598
DOI
EstadoPublicada - 2023
Evento24th International Conference on Digital Signal Processing, DSP 2023 - Rhodes, Grecia
Duración: 11 jun. 202313 jun. 2023

Serie de la publicación

NombreInternational Conference on Digital Signal Processing, DSP
Volumen2023-June

Conferencia

Conferencia24th International Conference on Digital Signal Processing, DSP 2023
País/TerritorioGrecia
CiudadRhodes
Período11/06/2313/06/23

Huella

Profundice en los temas de investigación de 'Enhancing safe screening rules with adaptive thresholding for non-overlapping group sparse norm regularized problems'. En conjunto forman una huella única.

Citar esto