taipeifx / Ames

ML techniques illustrated on Ames Housing Prices dataset.

The Machine Learning Project at NYCDSA

About

This group project uses Kaggle data on Ames (Iowa) housing prices and various machine learning techniques to build a predictive model.

Data

Besides original Kaggle data we add several more features:

• Dow Jones US Real Estate Index
• Corn prices
• Labor force in Ames
• Unemployment rate in Ames
• Fannie Mae mortgage rates.

All these variables are treated as lagged variables compareg to the date of house sale.

Models

We use 4 linear models, a Random Forest Regressor and an XGB model to make predictions.

Finally, we stack the models using the inverse of their error rate on a test set as weights.

The resulting RMSLE is 0.119