LJStroemsdoerfer/matcha: A parallelized two stage linear boosting algorithm

The package implements a self-designed Boosting algorithm. The procedure is based on two consecutive linear componentwise boosting runs, where features are choosen to minimize the MSE. The features returned after both boosting runs are then used to fit a penalized linear regression model.

Getting started

Package details

Authorperson("Lukas Jan", "Stroemsdoerfer", email = "[email protected]", role = c("aut", "cre"))
MaintainerLukas Stroemsdoerfer <[email protected]>
LicenseMIT
Version0.1
Package repositoryView on GitHub
Installation Install the latest version of this package by entering the following in R:
install.packages("devtools")
library(devtools)
install_github("LJStroemsdoerfer/matcha")
LJStroemsdoerfer/matcha documentation built on Aug. 30, 2018, 12:28 a.m.