On the Complexity of the Weighted Fussed Lasso

01/15/2018
by   José Bento, et al.
0

The solution path of the 1D fused lasso for an n-dimensional input is piecewise linear with O(n) segments (Hoefling et al. 2010 and Tibshirani et al 2011). However, existing proofs of this bound do not hold for the weighted fused lasso. At the same time, results for the generalized lasso, of which the weighted fused lasso is a special case, allow Ω(3^n) segments (Mairal et al. 2012). In this paper, we prove that the number of segments in the solution path of the weighted fused lasso is O(n^3), and for some instances Ω(n^2). We also give a new, very simple, proof of the O(n) bound for the fused lasso.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
01/15/2018

On the Complexity of the Weighted Fused Lasso

The solution path of the 1D fused lasso for an n-dimensional input is pi...
research
11/26/2014

A note relating ridge regression and OLS p-values to preconditioned sparse penalized regression

When the design matrix has orthonormal columns, "soft thresholding" the ...
research
03/29/2019

Variable selection and estimation in multivariate functional linear regression via the lasso

In more and more applications, a quantity of interest may depend on seve...
research
05/13/2020

Structure and Algorithm for Path of Solutions to a Class of Fused Lasso Problems

We study a class of fused lasso problems where the estimated parameters ...
research
06/14/2022

Graph-Based Spatial Segmentation of Health-Related Areal Data

Smoothing is often used to improve the readability and interpretability ...
research
05/13/2014

Efficient Implementations of the Generalized Lasso Dual Path Algorithm

We consider efficient implementations of the generalized lasso dual path...
research
10/27/2021

Denoising and change point localisation in piecewise-constant high-dimensional regression coefficients

We study the theoretical properties of the fused lasso procedure origina...

Please sign up or login with your details

Forgot password? Click here to reset