Predicting Real Estate Values Based on Open Data

Under market economy conditions, all companies are trying to gain advantage by securing market share, minimizing expenses and maximizing profits. The same is true for the real estate market and its short-term rental part. Significant part of this market belongs to peer-to-peer platform Airbnb, where homeowners can offer their property for rent for other platform users. The main goal of this work is to research machine learning methods application to the prediction of short-term rental prices and to create a price prediction model for Airbnb listings in order to give homeowners and renters a better understanding of rental pricing formation. Predicting the property price levels could help hosts to maximize their profits, and guests – to understand whether the accommodation is reasonably priced or not. In carticular, the role of the location and surrounding points of interest in the price formation is analysed.