Abstract:The application of the lasso is espoused in high-dimensional settings where only a small number of the regression coefficients are believed to be nonzero. Moreover, statistical properties of high-dimensional lasso estimators are often proved under the assumption that the correlation between the predictors is bounded. In this vein, coordinatewise methods, the most common means of computing the lasso solution, work well in the presence of low to moderate multicollinearity. The computational speed of coordinatewise algorithms degrades however as sparsity decreases and multicollinearity increases. Motivated by these limitations, we propose the novel "Deterministic Bayesian Lasso" algorithm for computing the lasso solution. This algorithm is developed by considering a limiting version of the Bayesian lasso. The performance of the Deterministic Bayesian Lasso improves as sparsity decreases and multicollinearity increases, and can offer substantial increases in computational speed. A rigorous theoretical analysis demonstrates that (1) the Deterministic Bayesian Lasso algorithm converges to the lasso solution, and (2) it leads to a representation of the lasso estimator which shows how it achieves both $\ell_1$ and $\ell_2$ types of shrinkage simultaneously. Connections to other algorithms are also provided. The benefits of the Deterministic Bayesian Lasso algorithm are then illustrated on simulated and real data.

Regression Coefficient and Autoregressive Order Shrinkage and Selection Via the Lasso

Average case analysis of Lasso under ultra-sparse conditions

Robust Regression Shrinkage and Consistent Variable Selection Through the LAD-Lasso

Regression shrinkage and selection via the lasso

Adaptive Lasso in high-dimensional settings

WLAD-LASSO Method for Robust Estimation and Variable Selection in Partially Linear Models

Lasso Regression: Estimation and Shrinkage via Limit of Gibbs Sampling

Regression selection and shrinkage via the lasso

Least Squares Approximation with a Diverging Number of Parameters

Lasso Adjustments Of Treatment Effect Estimates In Randomized Experiments

Information-Enriched Selection of Stationary and Non-Stationary Autoregressions using the Adaptive Lasso

On LASSO for predictive regression

Relaxed Adaptive Lasso and Its Asymptotic Results

Unified LASSO estimation by least squares approximation

Weighted LAD-LASSO method for robust parameter estimation and variable selection in regression

Robust Variable Selection Based on Relaxed Lad Lasso

The Adaptive Lasso and Its Oracle Properties

A Tuning-free Robust and Efficient Approach to High-dimensional Regression

Shrinkage Estimation of the Varying Coefficient Model

Canal-LASSO: A Sparse Noise-Resilient Online Linear Regression Model

Nonnegative Adaptive Lasso for Ultra-High Dimensional Regression Models and A Two-Stage Method Applied in Financial Modeling