# A critical review of LASSO and its derivatives for variable selection under dependence among covariates

arXiv: https://arxiv.org/abs/2012.11470 · HTML: https://mlsanity.com/paper/2012.11470/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 21 Dec 2020.

## Built on by (most cited first, 8 of 8)

- [Smoothing the Edges: Smooth Optimization for Sparse Regularization using Hadamard Overparametrization](https://mlsanity.com/paper/2307.03571/) (2023)
- [Continuous Semi-Supervised Nonnegative Matrix Factorization](https://mlsanity.com/paper/2212.09858/) (2022)
- [A Statistical View of Column Subset Selection](https://mlsanity.com/paper/2307.12892/) (2023)
- [Robust penalized least squares of depth trimmed residuals regression for high-dimensional data](https://mlsanity.com/paper/2309.01666/) (2023)
- [Structured Estimation of Heterogeneous Time Series](https://mlsanity.com/paper/2311.08658/) (2023)
- [Greedy feature selection: Classifier-dependent feature selection via greedy methods](https://mlsanity.com/paper/2403.05138/) (2024)
- [HIERVAR: A Hierarchical Feature Selection Method for Time Series Analysis](https://mlsanity.com/paper/2407.16048/) (2024)
- [Choosing the Right Regularizer for Applied ML: Simulation Benchmarks of Popular Scikit-learn Regularization Frameworks](https://mlsanity.com/paper/2604.03541/) (2026)

## Abstract (arXiv)

We study the limitations of the well known LASSO regression as a variable selector when there exists dependence structures among covariates. We analyze both the classic situation with n≥ p and the high dimensional framework with p>n. Restrictive properties of this methodology to guarantee optimality, as well as the inconveniences in practice, are analyzed. Examples of these drawbacks are showed by means of a extensive simulation study, making use of different dependence scenarios. In order to search for improvements, a broad comparison with LASSO derivatives and alternatives is carried out. Eventually, we give some guidance about what procedures are the best in terms of the data nature.