Introduction

This data set contains about 120k lines of data about hotel booking information.

Research question

• Predict if a hotel booking will be canceled using existing variables.

Files

data_dictionary.txt

This is a data dictionary, containing notes for each variable.

hotel_booking_prediction.ipynb

This is the jupyter notebook that contains all the code and analysis.

hotel_bookings.csv

This is the main data set used in this project.

Conclusions

Random forest does a better job at predicting cancellation, compared to logistic regression and decision tree, with 0.86 precision, 0.77 recall and 0.87 accuracy.