Penalized Regression in R

 You can copy and paste the recipes in this post to make a jump-start on your own problem or to learn and practice with linear regression in R.

Let’s get started.Each example in this post uses the longley dataset provided in the datasets package that comes with R. The longley dataset describes 7 economic variables observed from 1947 to 1962 used to predict the number of people employed yearly.

Ridge Regression

Ridge Regression creates a linear regression model that is penalized with the L2-norm which is the sum of the squared coefficients. This has the effect of shrinking the coefficient values (and the complexity of the model) allowing some coefficients with minor contribution to the response to get close to zero.

# load the package library(glmnet) # load data data(longley) x <- as.matrix(longley[,1:6]) y <- as.matrix(longley[,7]) # fit model fit <- glmnet(x, y, family="gaussian", alpha=0, lambda=0.001) # summarize the fit summary(fit) # make predictions predictions <- predict(fit, x, type="link") # summarize accuracy mse <- mean((y - predictions)^2) print(mse)

Learn about the glmnet function in the glmnet package.

Need more Help with R for Machine Learning?

Take my free 14-day email course and discover how to use R on your project (with sample code).

Click to sign-up and also get a free PDF Ebook version of the course.

Least Absolute Shrinkage and Selection Operator

Least Absolute Shrinkage and Selection Operator (LASSO) creates a regression model that is penalized with the L1-norm which is the sum of the absolute coefficients. This has the effect of shrinking coefficient values (and the complexity of the model), allowing some with a minor effect to the response to become zero.

# load the package library(lars) # load data data(longley) x <- as.matrix(longley[,1:6]) y <- as.matrix(longley[,7]) # fit model fit <- lars(x, y, type="lasso") # summarize the fit summary(fit) # select a step with a minimum error best_step <- fit$df[which.min(fit$RSS)] # make predictions predictions <- predict(fit, x, s=best_step, type="fit")$fit # summarize accuracy mse <- mean((y - predictions)^2) print(mse)

Learn about the lars function in the lars package.

Elastic Net

Elastic Net creates a regression model that is penalized with both the L1-norm and L2-norm. This has the effect of effectively shrinking coefficients (as in ridge regression) and setting some coefficients to zero (as in LASSO).

# load the package library(glmnet) # load data data(longley) x <- as.matrix(longley[,1:6]) y <- as.matrix(longley[,7]) # fit model fit <- glmnet(x, y, family="gaussian", alpha=0.5, lambda=0.001) # summarize the fit summary(fit) # make predictions predictions <- predict(fit, x, type="link") # summarize accuracy mse <- mean((y - predictions)^2) print(mse)

Learn about the glmnet function in the glmnet package.

Summary

In this post you discovered 3 recipes for penalized regression in R.

Penalization is a powerful method for attribute selection and improving the accuracy of predictive models. For more information see Chapter 6 of Applied Predictive Modeling by Kuhn and Johnson that provides an excellent introduction to linear regression with R for beginners.