Machine Learning in Hospitality: Interpretable Forecasting of Booking Cancellations

Abstract

The hospitality industry has a lot of problems with
money and how things run because people cancel their bookings
at the minute. This affects how money the hotels make and
how they plan their services. Now that we have a lot of booking
information we can use special computer programs to predict
when people will cancel.
This study is about making a computer program that can
predict when people will cancel their hotel bookings. We use
information about the bookings and the people who made them
to make predictions. We try out computer programs like Logistic
Regression and Decision Trees to see what makes people cancel.
We want to know why people cancel so we look at
things like how it takes for people to cancel what kind of
customers they are, if they paid a deposit and if they have
booked with us before. This helps hotel managers understand
when people might cancel and make plans to stop it from
happening. They can change their prices control how bookings
they take and talk to the customers who are most likely to cancel.
We test our computer program. It works well and is easy to
understand. This means that hotel managers can use it to make
decisions. Our results show that using these kinds of computer
programs can really help the hospitality industry make choices
based on data.
The hospitality industry can use this to make things better,
for themselves and their customers. Hotel booking cancellations
are a problem but with the right tools we can predict when they
will happen and do something about it. Hotel managers can use
this information to make their hotels better and make money.
The computer program is helpful because it can predict hotel
booking cancellations and help hotel managers understand why
people cancel.