Book Image

Machine Learning with R Cookbook, Second Edition - Second Edition

By : Yu-Wei, Chiu (David Chiu)
Book Image

Machine Learning with R Cookbook, Second Edition - Second Edition

By: Yu-Wei, Chiu (David Chiu)

Overview of this book

Big data has become a popular buzzword across many industries. An increasing number of people have been exposed to the term and are looking at how to leverage big data in their own businesses, to improve sales and profitability. However, collecting, aggregating, and visualizing data is just one part of the equation. Being able to extract useful information from data is another task, and a much more challenging one. Machine Learning with R Cookbook, Second Edition uses a practical approach to teach you how to perform machine learning with R. Each chapter is divided into several simple recipes. Through the step-by-step instructions provided in each recipe, you will be able to construct a predictive model by using a variety of machine learning packages. In this book, you will first learn to set up the R environment and use simple R commands to explore data. The next topic covers how to perform statistical analysis with machine learning analysis and assess created models, covered in detail later on in the book. You'll also learn how to integrate R and Hadoop to create a big data analysis platform. The detailed illustrations provide all the information required to start applying machine learning to individual projects. With Machine Learning with R Cookbook, machine learning has never been easier.
Table of Contents (21 chapters)
Title Page
Credits
About the Authors
About the Reviewers
www.PacktPub.com
Customer Feedback
Preface

Ranking the variable importance with the caret package


After building a supervised learning model, we can estimate the importance of features. This estimation employs a sensitivity analysis to measure the effect on the output of a given model when the inputs are varied. In this recipe, we will show you how to rank the variable importance with the caret package.

Getting ready

You need to have completed the previous recipe by storing the fitted rpart object in the model variable.

How to do it...

Perform the following steps to rank the variable importance with the caret package:

  1. First, you can estimate the variable importance with the varImp function:
    > importance = varImp(model, scale=FALSE)
    > importance
    Output
    rpart variable importance
    
                                  Overall
    number_customer_service_calls 116.015
    total_day_minutes             106.988
    total_day_charge              100.648
    international_planyes          86.789
    voice_mail_planyes   ...