AIRBNB Price Prediction Using Machine Learning

Airbnb is known to be a home-sharing and rental platform which provides facilities to homeowners or renters (referred to as hosts) to offer their houses otherwise known as listings on an online platform for guests booking. It is the responsibility of the hosts to set the expected price of their items independently. Although Airbnb along with a few sites provides many advices, we are yet to have any free or accurate system. This became difficult for the hosts to correctly come up with a price for their listed properties due to many different factors in the system. There are a few pricing models available in the market, however, these are not free. It is the responsibility of the host to enter the appropriate basic price for each night for a particular property. The other challenge is dynamic pricing based on holidays, seasons, and weather. The host can't keep the same price for all the dates as this impacts the business significantly. It is extremely critical to ensure appropriate prices are listed during this competitive time. This study compares the performance of numerous machine learning algorithms and methodologies in Airbnb price prediction to identify the most accurate one. Linear Regression, XGBoost, Random Forest, ANN and KNN are among the machine learning models experimented in this study. Different performance measures are used to validate the results.