Predicting hotel bookings
You just got a job at a hospitality research company, and your first task is to build a model that predicts whether or not a hotel stay will include children. To train your model, you will rely on a modified version of the hotel booking demands dataset by Antonio, Almeida, and Nunes (2019). You are restricting your data to the following features:
features <- c('hotel', 'adults',
'children', 'meal',
'reserved_room_type',
'customer_type',
'arrival_date')
The data has been loaded for you as hotels
, along with its corresponding test
and train
splits, and the model has been declared as lr_model <- logistic_reg()
.
You will assess model performance by accuracy and area under the ROC curve or AUC.
This exercise is part of the course
Feature Engineering in R
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
lr_recipe <-
recipe(children ~., data = train) %>%
# Generate "day of the week", "week" and "month" features
step_date(arrival_date, features = c(___, ___, ___)) %>%
# Create dummy variables for all nominal predictors
step_dummy(___)