Lava estimator
= Lava estimator
{c}
{title2=$\widehat\delta+\widehat\beta$}
= Regression with sparse and dense components
{synonym}
The Lava estimator minimizes $\|Y-X(\delta+\beta)\|_2^2/(2n)+\lambda_1\|\delta\|_1+\lambda_2\|\beta\|_2^2$. It combines a component selected by <Lasso> with a component shrunk by <ridge regression>. Profiling the latter yields a transformed <Lasso> problem; this is not the same coefficient penalty as <elastic net regularization>.