Machine Learning-Driven Analysis of Urban Hotel Performance: A Comparative Study

Machine Learning-Driven Analysis of Urban Hotel Performance: A Comparative Study of Manchester and European Benchmarks, by Dr. Nishanthi Rupika Abeynayake, Chiranjeewa Atapattu and Dr. Xia Cui
Abstract
This study investigates the performance of the urban hotel industry through a machine learning–driven forecasting framework, using hotel key performance indicators (KPIs) including Occupancy (OCC), Average Daily Rate (ADR), and Revenue per Available Room (RevPAR). The analysis is based on STR data from January 2018 to July 2025 across eight major European cities: Manchester, Amsterdam, Dublin, Lisbon, Madrid, Paris, Rome, and Vienna. To capture the complex dynamics of hotel performance, several statistical and machine learning models were employed, including ARIMA, SARIMA, Recurrent Neural Networks (RNN), Stochastic Gradient Boosting (XGBoots), Long Short-Term Memory (LSTM), and a hybrid ARIMA–XGBoost model. Model performance was evaluated using Mean Squared Error (MSE) and Mean Absolute Percentage Error (MAPE). Among these, the LSTM model consistently outperformed all other approaches, achieving forecasting accuracy with MAPE values below 5% across all cities. In addition to forecasting accuracy, seasonal decomposition was conducted for each city, revealing distinct patterns of seasonality in urban hotel markets. A focused weekly analysis of Manchester’s hotel industry provided insights into demand fluctuations and identified several strategic recommendations to enhance revenue management, pricing strategies, and capacity planning in the local hotel sector. The findings highlight the effectiveness of machine learning, particularly LSTM, in modelling hotel performance, and demonstrate the value of integrating advanced forecasting techniques to support data-driven decision-making in the hospitality industry.
Research Team
Dr. Nishanthi Rupika Abeynayake is a Senior Lecturer in Statistics and Data Science at Manchester Metropolitan University. She began her career as a Lecturer and rose to the position of Professor of Statistics at Wayamba University of Sri Lanka. In addition, she served as the President of the Institute of Applied Statistics, Sri Lanka, in 2016 and 2017. She holds a Ph.D. in Agricultural Statistics and has over 30 years of academic experience in teaching, research, and postgraduate supervision. Her research interests include time series forecasting, machine learning applications, and hospitality data analytics, with a focus on integrating statistical and AI-based methods for predictive modelling. Dr. Abeynayake has published widely in peer-reviewed journals and presented her work at numerous international conferences.
Chiranjeewa Atapattu is a Senior Lecturer and Faculty COIL Lead at Manchester Metropolitan University, with extensive experience in both the hospitality industry and higher education. His hospitality career included leadership roles at Hilton International and Millennium Hotels, where he specialized in operations, service excellence, and customer engagement. This industry expertise informs his academic pursuits, allowing him to incorporate practical insights into his teaching. Chiranjeewa holds a Master’s degree from York St John University and is a Fellow of the Higher Education Academy (UK). At Manchester Metropolitan, he lectures in Hospitality Services and Business Technology, focusing on Hospitality Business Analytics. He equips students to analyse data, interpret emerging trends, and make data driven decisions within a global service context. Beyond teaching, he has coached student teams for international business case competitions, achieving recognition in China, the Netherlands, and Egypt. His mentorship blends academic theory with practical application, supporting student achievement in competitive settings. With over 15 years of combined experience in both industry and academia, Chiranjeewa is dedicated to developing future leaders who can effectively bridge the gap between theory and practice in the global hospitality and service sectors.
Dr. Xia Cui is a Senior Lecturer in Software Engineering at the Manchester Metropolitan University and a member of the Human-Centred Computing Lab and Natural Language Processing (NLP) Lab at the Department of Computing and Mathematics. Previously, she completed a Knowledge Transfer Partnership (KTP) project to develop an AI powered system to detect speaker vulnerability indicators in telephone conversations with the University of Manchester and VoiceIQ Ltd. She received her PhD in Computer Science from the University of Liverpool in 2020, MSc in Web Science and Big Data Analytics from University College London in 2015, and BSc (Hons.) Internet Computing from the University of Liverpool in 2013. She has publications in major AI venues such as CVPR, InterSpeech, AACL, and IJCNN.