Predicting and explaining Airbnb prices in Lisbon : machine learning approach

• Autor(a) principal: Nunes, Madalena Ribeiro dos Santos Pais
• Data de Publicação: 2023
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10400.14/41431
• topic: Airbnb; Machine learning; Price prediction; xAI; Regression; Gradient boosting; Previsão de preços; Regressão; Domínio/Área Científica::Ciências Sociais::Economia e Gestão

Airbnb is an online platform that provides listing and arrangement for short-term local home renting services. Since its establishment in 2008, it has offered 7 million homes and rooms in more than 81,000 cities throughout 191 countries. Airbnb price prediction is a valuable and important task both for guests and hosts. Overall, for practical applications, these models can give a host an optimal price they should charge for their new listing. On the consumer side, this will help travellers determine whether the listing price they see is fair. Much research has been done in this field; however, the longitude and latitude of Airbnb listings are often disregarded. This project focuses on Airbnb price prediction using the most recent (Sep 2021) Airbnb data in Lisbon. Using Google Maps API, the original dataset was enriched with information on the number of ATMs, metro stations, bars and discos within a maximum radius of 1 km. Also, using the geodesic distance, the distance to the airport and the nearest attraction were computed for each listing. A Linear Regression and a Gradient Boosting algorithm were compared based on the original Airbnb dataset and the extended dataset to examine the impact of new features that have been identified. According to the results, all models perform better when the new features are included. The best results are achieved with the Gradient Boosting with the extended data, with an MAE of 0. 3102 and an adjusted R-squared of 0.4633.