11 Machine Learning & Predictive Analytics

11.0.1 Decision Trees

To prodcue decision tress you need to use variuos packages, such as rpart for recursive partioning and caret.

## Create a decision tree
library(rpart)
TreeResults <- rpart::rpart(force~mass + acceleration,
                     data=dataset1,
                     method="class",
                     control = rpart.control(minsplit = 30, cp = 0))

## Visualise the decision tree
rattle::fancyRpartPlot(TreeResults)

## Use model to predict value
predictions <- predict(TreeResults, dataset1, type = "class")

## Evaluate the results
confusionMatrix(predictions, dataset1$force)