Publication date: 03/23/2021

Image shown hereStatistical Details for Estimation Methods

Penalized regression methods introduce bias to the regression coefficients by penalizing them.

Image shown hereRidge Regression

An l2 penalty is applied to the regression coefficients during ridge regression. Ridge regression coefficient estimates are given by the following:

Equation shown here,

where Equation shown here is the l2 penalty, λ is the tuning parameter, N is the number of rows, and p is the number of variables.

Image shown hereDantzig Selector

An l penalty is applied to the regression coefficients during Dantzig Selector. Coefficient estimates for the Dantzig Selector satisfy the following criterion:

Equation shown here

where Equation shown here denotes the l norm, which is the maximum absolute value of the components of the vector v.

Image shown hereLasso Regression

An l1 penalty is applied to the regression coefficients during Lasso. Coefficient estimates for the Lasso are given by the following:

Equation shown here,

where Equation shown here is the l1 penalty, λ is the tuning parameter, N is the number of rows, and p is the number of variables

Image shown hereElastic Net

The Elastic Net combines both l1 and l2 penalties. Coefficient estimates for the Elastic Net are given by the following:

Equation shown here,

This is the notation used in the equation:

Equation shown here is the l1 penalty

Equation shown here is the l2 penalty

λ is the tuning parameter

α is a parameter that determines the mix of the l1 and l2 penalties

N is the number of rows

p is the number of variables

Tip: There are two sample scripts that illustrate the shrinkage effect of varying α and λ in the Elastic Net for a single predictor. Select Help > Sample Data, click Open the Sample Scripts Directory, and select demoElasticNetAlphaLambda.jsl or demoElasticNetAlphaLambda2.jsl. Each script contains a description of how to use it and what it illustrates.

Image shown hereAdaptive Methods

The adaptive Lasso method uses weighted penalties to provide consistent estimates of coefficients. The weighted form of the l1 penalty is

Equation shown here,

where Equation shown here is the MLE when the MLE exists. If the MLE does not exist and the response distribution is normal, estimation is done using least squares and Equation shown here is the solution obtained using a generalized inverse. If the response distribution is not normal, Equation shown here is the ridge solution.

For the adaptive Lasso, this weighted form of the l1 penalty is used in determining the Equation shown here coefficients.

The adaptive Elastic Net uses this weighted form of the l1 penalty and also imposes a weighted form of the l2 penalty. The weighted form of the l2 penalty for the adaptive Elastic Net is

Equation shown here,

where Equation shown here is the MLE when the MLE exists. If the MLE does not exist and the response distribution is normal, estimation is done using least squares and Equation shown here is the solution obtained using a generalized inverse. If the response distribution is not normal, Equation shown here is the ridge solution.

Want more information? Have questions? Get answers in the JMP User Community (community.jmp.com).