Director's Note

The AirBnB Price Predictor was a course project that aimed to help AirBnB hosts evaluate their own listing prices against their competition in New York City. The dataset was pulled from Kaggle and contained over 34,000 entries with 9 features each. With further feature engineering (splitting features into more useful features and scaling them) we were able to train and evaluate the performance of multiple machine learning models which included Linear Regression, K-Nearest Neighbour Regression, and Random Forest Regression. In the end, Random Forest Regression performed the best out of the three with the metrics of RMSE and MAPE scores. In the end, these models were also able to provide feature importance which revealed factors that hosts could improve on to improve their pricing and marketability.