Decision Trees

Download scoring and training data + packages

pacman::p_load(readr)
pacman::p_load(rpart)
pacman::p_load(rpart.plot)
pacman::p_load(rattle)
pacman::p_load(RColorBrewer)

ResumeNames_training <- read.csv("ResumeNames_training.csv")
ResumeNames_scoring <- read.csv("Resumenames_scoring.csv")

Build the decision tree model

call_back_tree <- rpart(call
                        ~gender+ethnicity+quality+holes+email+school+requirements
                            , data = ResumeNames_training
                            , method = "class"
                            , minsplit = 7
                            , minbucket = 1
                            , cp = -1
                           )

#1: if you wish to view your decision tree, plot it
#1: rpart.plot library
rpart.plot(call_back_tree, main = "Decision Tree for Callback on Resumes")

fancyRpartPlot(call_back_tree, caption = "Decision Tree for Callback on Resumes")

Scoring data - > make predictions based on the model

predictions <- predict(call_back_tree, ResumeNames_scoring)
View(predictions)

#1: join with scoring data

score_with_predict <- cbind(predictions, ResumeNames_scoring)
View(score_with_predict)