Μηχανική μάθηση στην πρόβλεψη της τιμής ενοικίασης Airbnb στο Άμστερνταμ
Machine learning prediction of Amsterdam Airbnb prices
In this thesis, the use and comparison of machine learning models to predict the price of an Airbnb listing is explored. More specifically, a dataset was collected from Airbnb listings in Amsterdam with a number of features such as property type, number of bedrooms, neighborhood, etc. The data was then cleaned and pre-processed to feed to the machine learning models. Several machine learning models including regression trees, random forest, and support vector machine were trained and tested on the dataset. The results of the different models were compared on the prediction accuracy by using MSE and R2 as metrics. The results showed that the XGBoost model had the highest predictive accuracy. However, the importance of features in terms of their contribution to making more accurate predictions was examined. Here the most important feature is proved that is the number of people who can stay in a property. Overall, the results of this study demonstrate the effectiveness of machine learning models in predicting the price of an Airbnb listing and highlights the importance of considering both the prediction accuracy and the influence of the most important features.