The hedonic price model of online short-term rental market based on machine learning

Abstract

Under the background of the sharing economy, the online short-term rental market contains huge business opportunities, but the development of online short-term rental models in various regions is uneven, and there is a lack of reasonable pricing. Therefore, based on the Beijing Airbnb online short-term rental data set, this paper adopts the advanced machine learning method AutoGluon model to predict the price range, and finally analyzes the solvability of the model. First, data preprocessing is performed on the initial dataset. Second, an initial exploration of the data found that prices are correlated with housing type, location, and surrounding environment. Then, based on the existing features, an AutoGluon hedonic price model is established to predict the price range; finally, interpretable analysis of the model is performed to identify key factors. Geographical location and room type undoubtedly have the greatest impact on the online short-term rental market price, providing a reference for online short-term rental platforms and homeowners to reasonably customize house prices and improve service quality.