Synopsis: Used linear regression techniques combined with feature selection and engineering to predict housing prices from same-market real estate listing data.
Methods: Lasso, Ridge, Feature Selection, pipelines, GridSearch
Data size: 1460 housing records with 79 features
Findings: Square footage, neighborhood primary predictors for price.