Inspiration
As rental rates continue to hit record highs in Singapore, our group is deeply interested in finding out the causal factors that determine rental prices in different districts.
Goal
We aim to help potential renters make better decisions in renting and assist the government in optimising the potential of a district through improved housing policies.
Solution
We based our model on several factors that we believed affects the rental rate the most, namely the size of the flat, its location, the number of lease years left as well as the number of amenities within a 2km radius. To build this model, we first web scraped data gov sg to procure data on the yearly rentals as well as the above factors. After securing the data, we cleaned and preprocessed the data to prepare it for training our model. Next, we split the data procured into train-test sets in a 7:3 ratio. Lastly, we used the data to create a polynomial regression model, in hopes of finding a casual relation between the factors and the rental price of HDBs in Singapore. At its current stage, our model is able to accurately predict the estimated rent of a HDB flat given its size, location, age and number of amenities within a 2km radius.
Challenges faced
A challenge we faced was incorporating the amenities factor into both past and present rental rates. As development in towns in Singapore takes place very rapidly, the number of amenities can change very drastically throughout the years. Hence, it was difficult for us to obtain data on amenities in the past and incorporate this change into our model.
Whats next for our project
Moving forward, we will be incorporating this into an interactive website that allows users to key in their desired factors (e.g location, number of rooms) to predict the rental rates. We also hope to implement a prediction system into our website, where buyers and sellers can agree on a preset price, and we would get a portion of the sales if a successful contract was made. We believe this service will both facilitate the rental market and transform our website into a lucrative business venture. This will also potentially increase the number of successful property contracts by matching buyers to sellers, which bolsters the housing market as a whole.
Log in or sign up for Devpost to join the conversation.