Talk:Elastic net regularization

Elastic Net After Rescaling

In the paper Regularization and variable selection via the elastic net by Hui Zou and Trevor Hastie, they state (in equation 14) the Elastic Net estimates after rescaling are given by:

${\hat {\beta }}=\arg \min \limits _{\beta }\beta ^{T}\left({\frac {\mathbf {X} ^{T}\mathbf {X} +\lambda _{2}\mathbf {I} }{1+\lambda _{2}}}\right)\beta -2\mathbf {y} ^{T}\mathbf {X} \beta +\lambda _{1}|\beta |_{1}$

This alternative form may be useful for some people. It allows one to optimize directly in the already rescaled space. It is also insightful as it provides insight as to why the authors rescaled by the factor $1+\lambda _{2}$. Mouse7mouse9 05:13, 27 October 2015 (UTC) — Preceding unsigned comment added by Mouse7mouse9 (talk • contribs)

Specification

In this section, we encounter the phrase "the authors" with no reference to whom is being referred to. If there's someone who feels particular ownership of this article, can you fix that? It's confusing and strange to not know who is being referred to. If you mean "the inventors" of elastic net, then refer that way and give their names (even if they're listed in the references). Thanks. Chafe66 (talk) 18:08, 15 June 2018 (UTC)