Learning-based Airbnb Price Prediction Model

Airbnb has become one of the largest online accommodation booking platforms today, providing more than 700 million accommodations in more than 220 countries. The main reason of successful booking is the proper price. To increase the success booking rate, the platform needs to provide price suggestions for the hosts. This paper focuses on Airbnb market of Beijing since China will be one of the main markets of Airbnb. The paper has developed a pricing prediction model based on machine learning approaches, i.e., XGBoost and neural network, for the Beijing Airbnb market. Through the analysis of the features related to the price, the research selects important ones to develop the prediction model. Apart from the accurate price prediction model, the research also gives suggestions for hosts about how to increase their price by adding important amenities.