Hotel Cancellation Prediction

[TOC]

Executive Summary

Our goal is to build a model able to classify a booking as canceled or not canceled. In order to do so, we used data from the Hotel Booking Demand Datasets. The dataset provides data from real bookings scheduled to arrive between July, 1st 2015 and August, 31st 2017 from two hotels in Portugal (a resort in the Algarve region (H1) and a hotel in the city of Lisbon (H2)). Booking data from both hotels share the same structure, with 31 variables describing the 40,060 observations of H1 and 79,330 observations of H2. For a detailed list and description of those variables refer to the data dictionary. The two hotel datasets were merged into one main dataframe. The dataframe was then cleaned making sure to address any null values, reformat certain features, and engineer new ones. Exploratory analysis included analysis of the cancellation target variable and its relation with other features. Data visualisation tools were used to identify trends and valuable insights from those analysis. A Random forest Classifier model was then used in order to identify the accuracy of the model. The models were evaluated, and conclusions and recommendations were derived to optimize occupancy, improve operations, and increase a hotel's revenue.

Quick Problem Overview

Who is our stakeholder? - Booking Manager
What is the core business problem we are solving? - Revenue Management, cost implications due to cancellations.
Convert Business to DS Problem: Determine if a booking will be canceled or not. And what leads to cancellations.
Business Metric: How much Overbookings can we take ( based on reliable cancelation predictions from the model)
Data Science Metric: Recall/ Precision/ F1 Score
Feature Selection: Done
Feature Engineering: Done
EDA Questions: Done

Problem with Cancellations

Customer accustomed to free cancellation policies
Operational Problems
Non accurate forecast
Non-optimized Occupancy
Poor Management
Revenue Loss

In order to fight the negative impact of cancellations, hotels need to be able to identify which bookings are likely to be cancelled.

Here the solution will allow hotels/resorts to predict if a new booking will be cancelled or not, manage their business accordingly, and increase their revenue.

The Data & Dictionary

From: Property Management System

Bookings are due to arrive between July 01, 2015 and August 31, 2017.

40,060 Hotel 1 (Algarve) | **79,330 **Hotel 2 (Lisbon) | 31 Variables

Feature Name	Type	Description
ADR	Float	Average Daily Rate. Calculated by dividing the sum of all lodging transactions by the total number of staying nights.
Adults	Integer	Number of adults.
Agent	Categorical	ID of the travel agency that made the booking.
ArrivalDateDayOfMonth	Integer	Day of the month of the arrival date.
ArrivalDateMonth	Categorical	Month of arrival date with 12 categories: “January” to “December”.
ArrivalDateWeekNumber	Integer	Week number of the arrival date.
ArrivalDateYear	Integer	Year of arrival date.
AssignedRoomType	Categorical	Code for the type of room assigned to the booking. Sometimes the assigned room type differs from the reserved room type due to hotel operation reasons (e.g. overbooking) or by customer request. Code is presented instead of designation for anonymity reasons.
Babies	Integer	Number of babies.
BookingChanges	Integer	Number of changes/amendments made to the booking from the moment the booking was entered on the Property Management System until the moment of check-in or cancellation. Calculated by adding the number of unique iterations that change some of the booking attributes, namely: persons, arrival date, nights, reserved room type or meal.
Children	Integer	Number of children. Sum of both payable and non-payable children.
Company	Categorical	ID of the company/entity that made the booking or responsible for paying the booking. ID is presented instead of designation for anonymity reasons.
Country	Categorical	Country of origin. Categories are represented in the International Standards Organization (ISO) 3155–3:2013 format.
CustomerType	Categorical	Type of booking, assuming one of four categories: Contract (when the booking has an allotment or other type of contract associated to it), Group (when the booking is associated to a group), Transient (when the booking is not part of a group or contract, and is not associated to other transient booking), and Transient-party (when the booking is transient, but is associated to at least other transient booking).
DaysInWaitingList	Integer	Number of days the booking was in the waiting list before it was confirmed to the customer. Calculated by subtracting the date the booking was confirmed to the customer from the date the booking entered on the Property Management System.
DepositType	Categorical	Indication on if the customer made a deposit to guarantee the booking. This variable can assume three categories: No Deposit (no deposit was made), Non Refund (a deposit was made in the value of the total stay cost), and Refundable (a deposit was made with a value under the total cost of stay). Value calculated based on the payments identified for the booking in the transaction (TR) table before the booking׳s arrival or cancellation date. In case no payments were found the value is “No Deposit”. If the payment was equal or exceeded the total cost of stay, the value is set as “Non Refund”. Otherwise the value is set as “Refundable”.
DistributionChannel	Categorical	Booking distribution channel. The term “TA” means “Travel Agents” and “TO” means “Tour Operators”.
IsCanceled	Integer	Value indicating if the booking was canceled (1) or not (0).
IsRepeatedGuest	Integer	Value indicating if the booking name was from a repeated guest (1) or not (0). Variable created by verifying if a profile was associated with the booking customer. If so, and if the customer profile creation date was prior to the creation date for the booking on the Property Management System database it was assumed the booking was from a repeated guest.
LeadTime	Integer	Number of days that elapsed between the entering date of the booking into the Property Management System and the arrival date. Calculated by subtracting the entering date from the arrival date.
MarketSegment	Categotical	Market segment designation. In categories, the term “TA” means “Travel Agents” and “TO” means “Tour Operators”.
Meals	Categorical	Type of meal booked. Categories are presented in standard hospitality meal packages: Undefined/SC (no meal package), BB (Bed & Breakfast), HB (Half board: breakfast and one other meal – usually dinner), and FB (Full board: breakfast, lunch and dinner).
PreviousBookingsNotCanceled	Integer	Number of previous bookings not canceled by the customer prior to the current booking. In case there was no customer profile associated with the booking, the value is set to 0. Otherwise, the value is the number of bookings with the same customer profile created before the current booking and not canceled.
PreviousCancellations	Integer	Number of previous bookings that were canceled by the customer prior to the current booking. In case there was no customer profile associated with the booking, the value is set to 0. Otherwise, the value is the number of bookings with the same customer profile created before the current booking and canceled.
RequiredCarParkingSpaces	Integer	Number of car parking spaces required by the customer.
ReservationStatus	Categorical	Reservation last status, assuming one of three categories: Canceled (booking was canceled by the customer), Check-Out (customer has checked in but already departed), No-Show (customer did not check-in and did inform the hotel of the reason why).
ReservationStatusDate	Date	Date at which the last status was set. This variable can be used in conjunction with the `ReservationStatus` to understand when was the booking canceled or when did the customer checked-out of the hotel.
ReservedRoomType	Categorical	Code of room type reserved. Code is presented instead of designation for anonymity reasons.
StaysInWeekendNights	Integer	Number of weekend nights (Saturday or Sunday) the guest stayed or booked to stay at the hotel. Calculated by counting the number of weekend nights from the total number of nights.
StaysInWeekNights	Integer	Number of week nights (Monday to Friday) the guest stayed or booked to stay at the hotel. Calculated by counting the number of week nights from the total number of nights.
TotalOfSpecialRequests	Integer	Number of special requests made by the customer (e.g. twin bed or high floor).

Workflow

Data Acquisition
Data Cleaning
EDA
Feature Engineering
Conclusion

Outcomes

Percentage of cancelled bookings overall is ~37%

Features correlated with cancellations

Lead Time
Previous Cancellations

Lead time booking per status

Days between booking and arrival

Cancelled bookings have longer average lead time
More time to cancel
More time for unexpected events

Special Requests

Canceled bookings have lower average number of special requests.
Engagement
Communication between customer and hotel

Parking Spaces

Cancelled bookings have lower average number of required parking spaces
Shows commitment to destination
Limit customer hotel opinion

Deposit Type

Customer who pay a non-refundable deposit have a much higher percentage of cancelled reservations
Transient groups who use a travel agent
Hotel deposit policies

manthangandhi / hotel_cancellation_prediction Goto Github PK