Hotel Reservations

Hotel Reservations

The hospitality industry faces a significant challenge with booking cancellations. When a guest cancels a reservation, hotels lose revenue and struggle to re-allocate rooms on short notice. High cancellation rates complicate resource planning, staffing, and inventory management.

You have been hired as a Data Scientist for a major hotel chain. Your task is to build a binary classification model to predict whether a booking will be canceled (is_canceled = True) or not. By identifying high-risk bookings, the hotel can implement strategic overbooking or targeted deposit policies.

Task

Using the provided historical data, train a model to predict the is_canceled status for the reservations in the test set. You are provided with:

• train.csv: Contains the features and the target labels (is_canceled) for training your model.
• test.csv: Contains the features for the reservations you need to predict. The target column is omitted.

Dataset

• id: Unique identifier for the reservation.
• hotel: The name/city of the hotel.
• lead_time: Number of days between the booking date and the arrival date.
• arrival_date_month: The month of arrival.
• deposit_type: Indicates if a deposit was made (No Deposit, Non Refund, Refundable).
• country: Country of origin of the guest.
• arrival_date_week_number: Week number of the year for arrival.
• stays_in_weekend_nights: Number of weekend nights (Saturday or Sunday) the guest stayed or booked to stay.
• stays_in_week_nights: Number of week nights (Monday to Friday) the guest stayed or booked to stay.
• is_repeated_guest: Indicator if the guest has stayed at the hotel before (0 or 1).
• required_car_parking_spaces: Number of car parking spaces required by the guest.
• adults/children/babies: Number of guests in each category.
• meal: Type of meal plan booked.
• booking_changes: Number of changes made to the booking.
• previous_cancellations: Number of previous bookings that were cancelled by the customer.
• total_of_special_requests: Number of special requests made by the guest (e.g. twin bed, high floor).
• customer_type: Type of booking (Contract, Group, Transient, etc.).
• resort_status: Boolean indicating if the property is a Resort Hotel.
• is_canceled: The target variable (True if canceled, False otherwise).

Evaluation

Submissions are evaluated based on the F1 Score.

• 100 points: F1 Score $\ge$ 0.75
• 0 to 99 points: F1 Score between 0.60 and 0.74 (calculated linearly).
• 0 points: F1 Score $<$ 0.60.

Submission Format

You must submit a CSV file with exactly two columns: id and answer. The answer column should contain your boolean predictions (True or False).

id,answer
1,False
2,True
3,False