On the differences between \(L_2\) boosting and the Lasso

DOI10.1016/j.spl.2019.108634zbMath1459.62137arXiv1812.05421OpenAlexW2976603259MaRDI QIDQ2288790

Michael Vogt

Publication date: 20 January 2020

Published in: Statistics \& Probability Letters (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/1812.05421

zbMATH Keywords

high-dimensional linear modelsrestricted eigenvalue condition\(L_2\) boostingparameter recovery/estimationrestricted nullspace property

Mathematics Subject Classification ID

Ridge regression; shrinkage estimators (Lasso) (62J07) Linear regression; mixed models (62J05)

Related Items

Estimation and inference of treatment effects with \(L_2\)-boosting in high-dimensional settings

Uses Software

• PDCO

Cites Work

• Unnamed Item
• Unnamed Item
• Greedy function approximation: A gradient boosting machine.
• Signal representation using adaptive normalized Gaussian functions
• Least angle regression. (With discussion)
• Weak greedy algorithms
• Simultaneous analysis of Lasso and Dantzig selector
• Forward stagewise regression and the monotone lasso
• The Dantzig selector: statistical estimation when \(p\) is much larger than \(n\). (With discussions and rejoinder).
• Boosting for high-dimensional linear models
• Compressed sensing and best 𝑘-term approximation
• Greedy Approximation
• Decoding by Linear Programming
• Stable recovery of sparse overcomplete representations in the presence of noise
• Sparse representations in unions of bases
• Greed is Good: Algorithmic Results for Sparse Approximation
• Atomic Decomposition by Basis Pursuit
• Uncertainty principles and ideal atomic decomposition
• On sparse representation in pairs of bases
• Matching pursuits with time-frequency dictionaries
• Optimally sparse representation in general (nonorthogonal) dictionaries via ℓ ¹ minimization