On the Closed-form Proximal Mapping and Efficient Algorithms for Exclusive Lasso Models

02/01/2019
by   Yancheng Yuan, et al.
0

The exclusive lasso regularization based on the ℓ_1,2 norm has become popular recently due to its superior performance over the group lasso regularization. Comparing to the group lasso regularization which enforces the competition on variables among different groups and results in inter-group sparsity, the exclusive lasso regularization also enforces the competition within each group and results in intra-group sparsity. However, to the best of our knowledge, a correct closed-form solution to the proximal mapping of the ℓ_1,2 norm has still been elusive. In this paper, we fill the gap by deriving a closed-form solution for Prox_ρ·_1^2(·) and its generalized Jacobian. Based on the obtained analytical results, we are able to design efficient first and second order algorithms for machine learning models involving the exclusive lasso regularization. Our analytical solution of the proximal mapping for the exclusive lasso regularization can be used to improve the efficiency of existing algorithms relying on the efficient computation of the proximal mapping.

READ FULL TEXT

page 16

page 24

page 25

research
08/17/2011

Structured Sparsity and Generalization

We present a data dependent generalization bound for a large class of re...
research
07/16/2021

Efficient proximal gradient algorithms for joint graphical lasso

We consider learning an undirected graphical model from sparse data. Whi...
research
04/16/2019

Discriminative Regression Machine: A Classifier for High-Dimensional Data or Imbalanced Data

We introduce a discriminative regression approach to supervised classifi...
research
06/16/2020

Efficient Path Algorithms for Clustered Lasso and OSCAR

In high dimensional regression, feature clustering by their effects on o...
research
06/02/2021

Smooth Bilevel Programming for Sparse Regularization

Iteratively reweighted least square (IRLS) is a popular approach to solv...
research
02/16/2022

Formulating Beurling LASSO for Source Separation via Proximal Gradient Iteration

Beurling LASSO generalizes the LASSO problem to finite Radon measures re...
research
08/30/2017

Graphical Lasso and Thresholding: Equivalence and Closed-form Solutions

Graphical Lasso (GL) is a popular method for learning the structure of a...

Please sign up or login with your details

Forgot password? Click here to reset