Publicação
Taking advantage of Data Science practices to optimise Revenue Management Strategies in the Hotel Industry: Development of a Price Recommendation Model
| Resumo: | With tourism continuing to grow every year, the hotel industry has become a key pillar of its success, generating significant revenue, with room bookings being one of the main contributors. Advances in technology have sparked greater interest in more sophisticated Revenue Management Systems. This dissertation focuses on developing a price recommendation model tailored to the unique needs of hotels in Portugal, aiming to predict optimal prices with accuracy. The dataset contained data from June 2018 until July 2021, containing 183812 observations and 22 variables. Before modelling, literature was conducted on pricing, forecasting and optimization algorithms for dynamic pricing tailored to the hotel industry. The data was tested using two approaches. In the first approach, we used normal regression models and in the second approach ensemble regression models. To ensure the model’s accuracy, the quality of forecasts was measured using Negative Mean Squared Error (MSE). To further improve our best-performing models, we performed a hyperparameterization on them so we could see if we could improve our results. The results show that ensemble models, particularly tree-based methods, are effective in predicting dynamic pricing. By fine-tuning these models with techniques like Grid Search and Optuna, their performance was further enhanced, leading to more precise and reliable price predictions. The research also revealed that factors like booking patterns and competitor pricing are key in determining optimal prices. This dissertation brings a developed perspective on how data science can enhance revenue management strategies in the hotel industry. By creating a price recommendation model that uses machine learning techniques, it moves beyond traditional methods and focuses on dynamic, data-driven pricing decisions. |
|---|---|
| Autores principais: | Silva, Maria Inês Alves Ferreira da |
| Assunto: | Pricing Optimization Machine Learning Revenue Maximisation Regressive Models SDG 8 - Decent work and economic growth SDG 9 - Industry, innovation and infrastructure |
| Ano: | 2025 |
| País: | Portugal |
| Tipo de documento: | dissertação de mestrado |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade Nova de Lisboa |
| Idioma: | inglês |
| Origem: | Repositório Institucional da UNL |
| Resumo: | With tourism continuing to grow every year, the hotel industry has become a key pillar of its success, generating significant revenue, with room bookings being one of the main contributors. Advances in technology have sparked greater interest in more sophisticated Revenue Management Systems. This dissertation focuses on developing a price recommendation model tailored to the unique needs of hotels in Portugal, aiming to predict optimal prices with accuracy. The dataset contained data from June 2018 until July 2021, containing 183812 observations and 22 variables. Before modelling, literature was conducted on pricing, forecasting and optimization algorithms for dynamic pricing tailored to the hotel industry. The data was tested using two approaches. In the first approach, we used normal regression models and in the second approach ensemble regression models. To ensure the model’s accuracy, the quality of forecasts was measured using Negative Mean Squared Error (MSE). To further improve our best-performing models, we performed a hyperparameterization on them so we could see if we could improve our results. The results show that ensemble models, particularly tree-based methods, are effective in predicting dynamic pricing. By fine-tuning these models with techniques like Grid Search and Optuna, their performance was further enhanced, leading to more precise and reliable price predictions. The research also revealed that factors like booking patterns and competitor pricing are key in determining optimal prices. This dissertation brings a developed perspective on how data science can enhance revenue management strategies in the hotel industry. By creating a price recommendation model that uses machine learning techniques, it moves beyond traditional methods and focuses on dynamic, data-driven pricing decisions. |
|---|