Parallelizable algorithms for the selection of grouped variables

Gonzalo Mateos, Juan Andrés Bazerque, Georgios B. Giannakis

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Scopus citations

Abstract

Well-appreciated in statistics for its ability to select relevant grouped features (factors) in linear regression models, the group-Lasso estimator has been fruitfully applied to diverse signal processing problems including RF spectrum cartography and robust layered sensing. These applications motivate the distributed group-Lasso algorithm developed in this paper, that can be run by a network of wireless sensors, or, by multiple processors to balance the load of a single computational unit. After reformulating the group-Lasso cost into a separable form, it is iteratively minimized using the method of multipliers to obtain parallel per agent and per factor estimate updates given by vector soft-thresholding operations. Through affordable inter-agent communication of sparse messages, the local estimates provably consent to the global group-Lasso solution. Specializing to a single agent network, or, to univariate factors, efficient (distributed) Lasso solvers are rediscovered as a byproduct.

Original languageEnglish (US)
Title of host publication2011 Digital Signal Processing and Signal Processing Education Meeting, DSP/SPE 2011 - Proceedings
Pages295-300
Number of pages6
DOIs
StatePublished - Apr 21 2011
Event2011 Digital Signal Processing and Signal Processing Education Meeting, DSP/SPE 2011 - Sedona, AZ, United States
Duration: Jan 4 2011Jan 7 2011

Publication series

Name2011 Digital Signal Processing and Signal Processing Education Meeting, DSP/SPE 2011 - Proceedings

Other

Other2011 Digital Signal Processing and Signal Processing Education Meeting, DSP/SPE 2011
CountryUnited States
CitySedona, AZ
Period1/4/111/7/11

Keywords

  • (group-) Lasso
  • Sparsity
  • distributed estimation
  • linear regression
  • parallel optimization

Fingerprint Dive into the research topics of 'Parallelizable algorithms for the selection of grouped variables'. Together they form a unique fingerprint.

Cite this