MACHINE LEARNING APPROACH TO PREDICT HOTEL RESERVATION CANCELLATIONS

Abstract

Hotel reservation cancellations present a critical challenge in the hospitality industry,
resulting in revenue losses, inefficient resource utilization, and operational instability.
Accurately predicting cancellations can help hotels improve occupancy rates,
optimize revenue management strategies, and enhance customer satisfaction. This
study proposes a machine learning-based approach to predict hotel reservation
cancellations by analyzing guest booking patterns and lead-time behavior.
The dataset used in this study comprises historical hotel booking records, including
features such as lead time, number of guests (adults and children), booking channel,
market segment, room type, special requests, and previous cancellation history. The
target variable indicates whether a reservation is canceled or not. Data preprocessing
techniques such as handling missing values, encoding categorical variables, and
feature scaling were applied to ensure data quality. Exploratory Data Analysis (EDA)
was conducted to identify key patterns and relationships, revealing that factors like
longer lead times, prior cancellations, and certain market segments significantly
influence cancellation likelihood.
Machine learning models including Decision Tree, Random Forest, and Gradient
Boosting classifiers were implemented to build the predictive system. Model
performance was evaluated using metrics such as accuracy, precision, recall, F1-score,
and ROC-AUC. Additionally, feature importance analysis was performed to identify
the most influential factors affecting cancellations. To enhance usability, a Streamlitbased
dashboard was developed, enabling hotel managers to visualize predictions and
make informed decisions