Airbnb is an online platform that enables property owners to put their vacancy room online so that guests can find an accomodation. In order to maximise the profit for the property owner and find a best deal for the guest, setting a price plays an important role in their retention in the platform. In this case study, I am going to use the dataset by Airbnb to predict the listing price based upon the feature.

(1) How does geographical location affect the price and How much does it contribute to the model?
(2) How does the room quality affect the price and How much does it contribute to the model?
(3) How does text data affect the listing price and how much does it contibute to the model

Outlier reduction (IQR measure)

Spatial Visualisation

Natural Language Processing

Feature Engineering based upon domain knowledge

Random forest with K-fold Random Search

Permutative future importance

RMSE