%0 Journal Article
%T Gather and Conquer: Region-Based Strategies to Accelerate Safe Screening Tests
%+ SIMulation pARTiculaire de Modèles Stochastiques (SIMSMART)
%+ Lab-STICC_ENSTAB_CID_TOMS
%A Herzet, Cedric
%A Dorffer, Clément
%A Drémeau, Angélique
%Z DGA/MRIS
%Z N62909-17-1-2007, ONR
%Z ANR-15-CE23-0021, ANR
%Z CominLabs
%< avec comité de lecture
%@ 1053-587X
%J IEEE Transactions on Signal Processing
%I Institute of Electrical and Electronics Engineers
%V 67
%N 12
%P 3300-3315
%8 2019-06-15
%D 2019
%R 10.1109/TSP.2019.2914885
%K Computational complexity
%K Regression analysis
%K Statistical testing
%Z Computer Science [cs]/Signal and Image Processing
%Z Mathematics [math]/Optimization and Control [math.OC]
%Z Statistics [stat]/Machine Learning [stat.ML]Journal articles
%X In this paper, we propose new methodologies to decrease the computational cost of safe screening tests for LASSO. We first introduce a new screening strategy, dubbed “joint screening test,” which allows the rejection of a set of atoms by performing one single test. Our approach enables to find good compromises between complexity of implementation and effectiveness of screening. Second, we propose two new methods to decrease the computational cost inherent to the construction of the (so-called) “safe region.” Our numerical experiments show that the proposed procedures lead to significant computational gains as compared to standard methodologies.
%G English
%L hal-02303198
%U https://hal-ensta-bretagne.archives-ouvertes.fr/hal-02303198
%~ UNIV-BREST
%~ INSTITUT-TELECOM
%~ ENSTA-BRETAGNE
%~ UNIV-RENNES1
%~ IRMAR
%~ UR2-HB
%~ CNRS
%~ INRIA
%~ UNIV-UBS
%~ INSA-RENNES
%~ INRIA-RENNES
%~ ENSTA-BRETAGNE-STIC
%~ INSMI
%~ INRIA_TEST
%~ UNAM
%~ TESTALAIN1
%~ ENIB
%~ LAB-STICC
%~ CHL
%~ INRIA2
%~ TDS-MACS
%~ UR1-HAL
%~ UR1-MATH-STIC
%~ AGREENIUM
%~ UNIV-RENNES2
%~ TEST-UR-CSS
%~ UNIV-RENNES
%~ INRIA-RENGRE
%~ INSA-GROUPE
%~ INSTITUTS-TELECOM
%~ TEST-HALCNRS
%~ ANR
%~ UR1-MATH-NUM
%~ INRIAARTDOI
%~ INSTITUT-AGRO