#Perform Naive Bayes on the Iris Data Set # # Author: mike-bowles PatriciaHoffman ############################################################################### rm(list=ls()) #install.packages("klaR") #install.packages("e1071") library(klaR) require(klaR) data(iris) #have a look at the data iris labels(iris) m <- NaiveBayes(Species ~ ., data = iris) #have a look at posterior densities as a function of individual variables data(iris) mN <- NaiveBayes(Species ~ ., data = iris) str(mN) plot(mN) #Here's a visualization of the decision boundaries in two dimensions library(MASS) data(iris) partimat(Species ~ ., data = iris, method = "naiveBayes") #what does a prediction look like? data(iris) m <- NaiveBayes(Species ~ ., data = iris) out <- predict(m) Err <- 1 - sum(out$class == iris$Species)/length(iris$Species) Err