Machine Learning with R Cookbook

Machine Learning with R Cookbook

By : Yu-Wei, Chiu (David Chiu)

Buy this Book

Machine Learning with R Cookbook

By: Yu-Wei, Chiu (David Chiu)

Buy this Book

Overview of this book

<p>The R language is a powerful open source functional programming language. At its core, R is a statistical programming language that provides impressive tools to analyze data and create high-level graphics.</p> <p>This book covers the basics of R by setting up a user-friendly programming environment and performing data ETL in R. Data exploration examples are provided that demonstrate how powerful data visualization and machine learning is in discovering hidden relationships. You will then dive into important machine learning topics, including data classification, regression, clustering, association rule mining, and dimension reduction.</p>

Machine Learning with R Cookbook

Credits

About the Author

About the Reviewers

www.PacktPub.com

Preface

Free Chapter

Practical Machine Learning with R

Introduction

Downloading and installing R

Downloading and installing RStudio

Installing and loading packages

Reading and writing data

Using R to manipulate data

Applying basic statistics

Visualizing data

Getting a dataset for machine learning

Data Exploration with RMS Titanic

Introduction

Reading a Titanic dataset from a CSV file

Converting types on character variables

Detecting missing values

Imputing missing values

Exploring and visualizing data

Predicting passenger survival with a decision tree

Validating the power of prediction with a confusion matrix

Assessing performance with the ROC curve

R and Statistics

Introduction

Understanding data sampling in R

Operating a probability distribution in R

Working with univariate descriptive statistics in R

Performing correlations and multivariate analysis

Operating linear regression and multivariate analysis

Conducting an exact binomial test

Performing student's t-test

Performing the Kolmogorov-Smirnov test

Understanding the Wilcoxon Rank Sum and Signed Rank test

Working with Pearson's Chi-squared test

Conducting a one-way ANOVA

Performing a two-way ANOVA

Understanding Regression Analysis

Introduction

Fitting a linear regression model with lm

Summarizing linear model fits

Using linear regression to predict unknown values

Generating a diagnostic plot of a fitted model

Fitting a polynomial regression model with lm

Fitting a robust linear regression model with rlm

Studying a case of linear regression on SLID data

Applying the Gaussian model for generalized linear regression

Applying the Poisson model for generalized linear regression

Applying the Binomial model for generalized linear regression

Fitting a generalized additive model to data

Visualizing a generalized additive model

Diagnosing a generalized additive model

Classification (I) – Tree, Lazy, and Probabilistic

Introduction

Preparing the training and testing datasets

Building a classification model with recursive partitioning trees

Visualizing a recursive partitioning tree

Measuring the prediction performance of a recursive partitioning tree

Pruning a recursive partitioning tree

Building a classification model with a conditional inference tree

Visualizing a conditional inference tree

Measuring the prediction performance of a conditional inference tree

Classifying data with the k-nearest neighbor classifier

Classifying data with logistic regression

Classifying data with the Naïve Bayes classifier

Classification (II) – Neural Network and SVM

Introduction

Classifying data with a support vector machine

Choosing the cost of a support vector machine

Visualizing an SVM fit

Predicting labels based on a model trained by a support vector machine

Tuning a support vector machine

Training a neural network with neuralnet

Visualizing a neural network trained by neuralnet

Predicting labels based on a model trained by neuralnet

Training a neural network with nnet

Predicting labels based on a model trained by nnet

Model Evaluation

Introduction

Estimating model performance with k-fold cross-validation

Performing cross-validation with the e1071 package

Performing cross-validation with the caret package

Ranking the variable importance with the caret package

Ranking the variable importance with the rminer package

Finding highly correlated features with the caret package

Selecting features using the caret package

Measuring the performance of the regression model

Measuring prediction performance with a confusion matrix

Measuring prediction performance using ROCR

Comparing an ROC curve using the caret package

Measuring performance differences between models with the caret package

Ensemble Learning

Introduction

Classifying data with the bagging method

Performing cross-validation with the bagging method

Classifying data with the boosting method

Performing cross-validation with the boosting method

Classifying data with gradient boosting

Calculating the margins of a classifier

Calculating the error evolution of the ensemble method

Classifying data with random forest

Estimating the prediction errors of different classifiers

Clustering

Introduction

Clustering data with hierarchical clustering

Cutting trees into clusters

Clustering data with the k-means method

Drawing a bivariate cluster plot

Comparing clustering methods

Extracting silhouette information from clustering

Obtaining the optimum number of clusters for k-means

Clustering data with the density-based method

Clustering data with the model-based method

Visualizing a dissimilarity matrix

Validating clusters externally

Association Analysis and Sequence Mining

Introduction

Transforming data into transactions

Displaying transactions and associations

Mining associations with the Apriori rule

Pruning redundant rules

Visualizing association rules

Mining frequent itemsets with Eclat

Creating transactions with temporal information

Mining frequent sequential patterns with cSPADE

Dimension Reduction

Introduction

Performing feature selection with FSelector

Performing dimension reduction with PCA

Determining the number of principal components using the scree test

Determining the number of principal components using the Kaiser method

Visualizing multivariate data using biplot

Performing dimension reduction with MDS

Reducing dimensions with SVD

Compressing images with SVD

Performing nonlinear dimension reduction with ISOMAP

Performing nonlinear dimension reduction with Local Linear Embedding

Big Data Analysis (R and Hadoop)

Introduction

Preparing the RHadoop environment

Installing rmr2

Installing rhdfs

Operating HDFS with rhdfs

Implementing a word count problem with RHadoop

Comparing the performance between an R MapReduce program and a standard R program

Testing and debugging the rmr2 program

Installing plyrmr

Manipulating data with plyrmr

Conducting machine learning with RHadoop

Configuring RHadoop clusters on Amazon EMR

Resources for R and Machine Learning

Dataset – Survival of Passengers on the Titanic

Index

Customer Reviews

5 star

4 star

3 star

2 star

1 star

Performing cross-validation with the e1071 package

Besides implementing a loop function to perform the k-fold cross-validation, you can use the tuning function (for example, tune.nnet, tune.randomForest, tune.rpart, tune.svm, and tune.knn.) within the e1071 package to obtain the minimum error value. In this recipe, we will illustrate how to use tune.svm to perform the 10-fold cross-validation and obtain the optimum classification model.

Getting ready

In this recipe, we continue to use the telecom churn dataset as the input data source to perform 10-fold cross-validation.

How to do it...

Perform the following steps to retrieve the minimum estimation error using cross-validation:

Apply tune.svm on the training dataset, trainset, with the 10-fold cross-validation as the tuning control. (If you find an error message, such as could not find function predict.func, please clear the workspace, restart the R session and reload the e1071 library again):
```
> tuned = tune.svm(churn~., data = trainset, gamma = 10^-2, cost = 10^2, tunecontrol=tune.control(cross=10))
```

Next, you can obtain the summary information of the model, tuned:

> summary(tuned)

Error estimation of 'svm' using 10-fold cross validation: 0.08164651

Then, you can access the performance details of the tuned model:

> tuned$performances
  gamma cost      error dispersion
1  0.01  100 0.08164651 0.02437228

Lastly, you can use the optimum model to generate a classification table:

> svmfit = tuned$best.model
> table(trainset[,c("churn")], predict(svmfit))
     
       yes   no
  yes  234  108
  no    13 1960

How it works...

The e1071 package provides miscellaneous functions to build and assess models, therefore, you do not need to reinvent the wheel to evaluate a fitted model. In this recipe, we use the tune.svm function to tune the svm model with the given formula, dataset, gamma, cost, and control functions. Within the tune.control options, we configure the option as cross=10, which performs a 10-fold cross validation during the tuning process. The tuning process will eventually return the minimum estimation error, performance detail, and the best model during the tuning process. Therefore, we can obtain the performance measures of the tuning and further use the optimum model to generate a classification table.

Machine Learning with R Cookbook

By : Yu-Wei, Chiu (David Chiu)

Machine Learning with R Cookbook

By: Yu-Wei, Chiu (David Chiu)

Overview of this book

Related Content you might be interested in

Current Title:

Machine Learning with R Cookbook

Performing cross-validation with the e1071 package

Getting ready

How to do it...

How it works...

See also