Book Image

Learning Predictive Analytics with R

By : Eric Mayor
Book Image

Learning Predictive Analytics with R

By: Eric Mayor

Overview of this book

This book is packed with easy-to-follow guidelines that explain the workings of the many key data mining tools of R, which are used to discover knowledge from your data. You will learn how to perform key predictive analytics tasks using R, such as train and test predictive models for classification and regression tasks, score new data sets and so on. All chapters will guide you in acquiring the skills in a practical way. Most chapters also include a theoretical introduction that will sharpen your understanding of the subject matter and invite you to go further. The book familiarizes you with the most common data mining tools of R, such as k-means, hierarchical regression, linear regression, association rules, principal component analysis, multilevel modeling, k-NN, Naïve Bayes, decision trees, and text mining. It also provides a description of visualization techniques using the basic visualization tools of R as well as lattice for visualizing patterns in data organized in groups. This book is invaluable for anyone fascinated by the data mining opportunities offered by GNU R and its packages.
Table of Contents (23 chapters)
Learning Predictive Analytics with R
Credits
About the Author
About the Reviewers
www.PacktPub.com
Preface
Exercises and Solutions
Index

Working with multiple regression


In multiple regression, we are interested in testing the impact of several predictors on a criterion, instead of just one in simple regression. Here, the value of the observations can be computed as the intercept plus the slope coefficient multiplied by the predictor value (for each predictor) plus the residuals.

The analysis estimates the unique contribution of the predictors to the criterion—that is, each obtained slope coefficient value (there is one for each predictor) and the intercepts that are controlled for the influence of the other predictors on the criterion. We are not going to detail the calculation of the slope and intercept for multiple regression as this involves more complex explanations than for simple regression and will not add much to your understanding; most of what we have seen (except the calculation of the coefficients and degrees of freedom) remains valid for multiple regression. We will now directly skip to a more practical section...