Naive Bayes originally became popular as a method for spam detection. It is quick and fast. Naive Bayes assumes that the variables are all independent and not related to each other (a bad assumption, but that is what makes it Naive!). It also has the advantage that it does not need to be retrained when adding new data. Naive Bayes has its roots in Bayes theorem.
This simple example shows Naive Bayes in action. Using the Iris dataset, Naive Bayes will make a prediction for the fifth column using the first four columns as independent variables:
#use 1st 4 columns to predict the fifthlibrary(e1071)iris.nb<-naiveBayes(iris[,1:4], iris[,5])table(predict(iris.nb, iris[,1:4]), iris[,5])
The results of the table() function ...