The Data Analysis Tools Series at the Berkeley Institute for Data Science

Ensemble (Machine) Learning with Super Learner and H2O in R -- Nima Hejazi and Evan Muzzall

December 6, 2016 at 5-6:30pm in BIDS, 190 Doe Library

Nima Hejazi & Evan Muzzall

Nima is a graduate student in the Division of Biostatistics. His research combines aspects of causal inference, statistical machine learning, and nonparametric statistics, with a focus on the development of robust methods for addressing inference problems arising in precision medicine, computational biology, and clinical trials.

Evan earned his Ph.D. in Biological Anthropology from Southern Illinois University Carbondale where he focused on spatial patterns of skeletal and dental variation in two large necropoles of Iron Age Central Italy (1st millennium BC). He is currently R Lead Instructor, co-founder of the Machine Learning Working Group, and Research Associate in the D-Lab.

Ensemble (Machine) Learning with Super Learner and H2O in R

This presentation covers methods for performing ensemble machine learning with the Super Learner R package and H2O software platform, using the R language for statistical computing.

Materials for this presentation are available on GitHub here.

R & RStudio Installation

Jupyter R Kernel Installation

SuperLearner Installation

require("devtools")
devtools::install_github("ecpolley/SuperLearner")

H2O Installation

These installations are required to make H2O work in RStudio. Click the links to visit the download pages.

  1. Download RStudio

  2. Download Java Runtime Environment

  3. Download H2O for R and dependencies (click the “Use H2O directly from R” tab and follow the copy/paste instructions)

  4. Install the devtools and h2oEnsemble R packages.

# The following two commands remove any previously installed H2O packages for R.
if ("package:h2o" %in% search()) { detach("package:h2o", unload=TRUE) }
if ("h2o" %in% rownames(installed.packages())) { remove.packages("h2o") }

# Next, we download packages that H2O depends on.
pkgs <- c("methods","statmod","stats","graphics","RCurl","jsonlite","tools","utils")
for (pkg in pkgs) {
if (! (pkg %in% rownames(installed.packages()))) { install.packages(pkg, repos = "http://cran.rstudio.com/") }
}

# Now we download, install and call the H2O package for R.
install.packages("h2o", type="source", repos=(c("http://h2o-release.s3.amazonaws.com/h2o/rel-turing/10/R")))

# Install the "devtools" R package.
install.packages(c("devtools"))

# Install the "h2oEnsemble" R package.
install_github("h2oai/h2o-3/h2o-r/ensemble/h2oEnsemble-package")

# Load packages
library(h2o)
library(devtools)
library(h2oEnsemble)

Lightning Talks

<+ person +> : <+ topic +>

<+ person +> : <+ topic +>

Share