Ensembling Regularized Linear Models

Functions for computing the ensembles of regularized linear regression estimators defined in Christidis, Lakshmanan, Smucler and Zamar (2017) . The procedure works on top of a given penalized linear regression estimator, the Elastic Net in this implementation, by fitting it to possibly overlapping subsets of features, while at the same time encouraging diversity among the subsets, to reduce the correlations between the predictions that result from each fitted model. The predictions from the models are then aggregated.


Build Status CRAN_Status_Badge Downloads

This package provides functions for computing the ensembles of regularized linear regression estimators defined in Christidis, Lakshmanan, Smucler and Zamar (2017).


You can install the stable version on R CRAN.

install.packages('ensembleEN', dependencies = TRUE)

You can install the development version from GitHub

library(devtools)
devtools::install_github("esmucler/ensembleEN")

Usage

# A small example
library(MASS)
library(ensembleEN)
set.seed(1)
beta <- c(rep(5, 5), rep(0, 45))
Sigma <- matrix(0.5, 50, 50)
diag(Sigma) <- 1
x <- mvrnorm(50, mu = rep(0, 50), Sigma = Sigma)
y <- x %*% beta + rnorm(50)
fit <- cv.ensembleEN(x, y, num_models=10) # Use 10 models
coefs <- predict(fit, type="coefficients")

License

This package is free and open source software, licensed under GPL (>= 2).

News

ensembleEN 1.1.2

  • Add new test, fix old ones, specially the objective function test.
  • Decrease default tolerance.

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.

install.packages("ensembleEN")

1.1.2 by Ezequiel Smucler, a year ago


Browse source code at https://github.com/cran/ensembleEN


Authors: Anthony Christidis <[email protected]> , Ezequiel Smucler <[email protected]> , Ruben Zamar <[email protected]>


Documentation:   PDF Manual  


GPL (>= 2) license


Imports Rcpp

Suggests testthat, glmnet, MASS

Linking to Rcpp, RcppArmadillo


See at CRAN