Ames Housing Price Prediction App

Competition Description

Ask a home buyer to describe their dream house, and they probably won't begin with the height of the basement ceiling or the proximity to an east-west railroad. But this playground competition's dataset proves that much more influences price negotiations than the number of bedrooms or a white-picket fence.

With 79 explanatory variables describing (almost) every aspect of residential homes in Ames, Iowa, this competition challenges you to predict the final price of each home.

Summary of visualization

  1. 1-saleprice and overall quality is strongly corelated.
  2. 2-garagearea and garagecars(Size of garage in square feet & Size of garage in car capacity)
  3. 3-totalbsmtSF and 1stFlrSF(Total square feet of basement area & First Floor square feet)
  4. 4-GrLivArea and TotRmsAbvGrd(Above grade (ground) living area square feet & Total rooms above grade (does not include bathrooms))

After initial analysis I have trained different models and selected the one with overall best performance. In this competition we are using MSE as the measure of error.

I have compared the following models:

  1. Linear Regression
  2. Lasso
  3. Ridge
  4. Decision Tree
  5. Random Forest

Ridge regression model came out on top with lowest MSE on test set.

Leaderboard

The App

After finalizing the model the next set was to build a front-end through which a user can give in inputs to the model and receive output generated by the model. To buid a responsive front-end I have used the following tech stack.

There was need of an API endpoint which will feed our inputs from front-end to the model and deliver the result back to out UI. Flask was used for the task.

After everything was responding perfectly. I started a Unix instance on AWS EC2 and with the help of NGINX server deployed my model on the cloud.

RESULT