chapter  16
14 Pages

R Packages for Neural Networks, Deep Learning, and Naïve Bayes

ByWilliam B. Claster

Almost every machine learning algorithm used in practice is available as a package in R. One package called caret includes many different algorithms all under one unified set of commands, and so once we learn how to make predictions with one of the algorithms within caret, the interfaces are all similar within caret. This chapter looks into Naive Bayes and neural network packages. The Naive Bayes algorithm falls under the category of multinomial Naive Bayes algorithms. The chapter explores two Naive Bayes algorithm packages. One comes from a package with the somewhat off-putting name of e1071. The other package explores is called naivebayes. The chapter explores the neural network package called neuralnet. Scaling is often performed to map the entire range of data for a particular attribute to either of the intervals.