BLACK FRIDAY EXTRA SAVINGS EVENT - EXTENDED
START FREE

ML rages on - house prices project

Y’all know what this is: linear regression model for predicting house prices, here’s whats inside:

  • the usual project scope and…
  • a big chunk of feature engineering (halved the amount of columns in the dataset)
  • outlier removal basics
  • overfitting / underfitting - used a validation test at the end
  • started considering memory constrains (not a lot of that issue in this notebook, but started a separate project for that matter)

project on Github

houses_final.ipynb (836.0 KB)

Things to consider in the future:

  • use more sklearn (SelectKBest, Imputer etc.)
  • np.log transformation
  • use of k-fold for validation

Any feedback welcome!

Click here to view the jupyter notebook file in a new tab

Last screen link:
https://www.dataquest.io/c/41/m/240/guided-project%3A-predicting-house-sale-prices

2 Likes