Inspiration
Our inspiration came from a Kaggle competition that tasked us with predicting Airbnb rental prices in New York City.
What it does
Our model creates a prediction for the sales price of houses in three states: Massachusetts, Rhode Island, and Pennsylvania.
How we built it
We used a gradient boosting algorithm to take into consideration different features both from the dataset we were given and some that we feature engineered ourselves.
Challenges we ran into
We were in a constant battle against time. Running prediction models can be time-consuming, and we spent most of our time feature engineering variables while we could have paid more attention to the tuning of our model.
Accomplishments that we're proud of
There are many things that we will bring home from this experience: an absolute sense of confidence in our capabilities, the awareness that we have the grit to succeed and the satisfaction of having made it to the end of the challenge in the best mood and spirit.
What we learned
We learned that no task is too large for our group!
What's next for House Prices Prediction
An in-depth tuning session! There is a lot to be done and we are excited to get our model up and running to minimize our RMSE.
Log in or sign up for Devpost to join the conversation.