American Housing Stock

Classifying Adequacy of Units Based on Neighborhood Characteristics

Contributors:

• Seoho Hahm (@seoho926)
• Nicole Bills (@NicoleJaneway)

Goal:

The goal of this analysis is to assist U.S. Department of Housing and Urban Development (HUD) to better understand the adequacy of housing units based on neighborhood characteristics.

Methods Used:

• Supervised Machine Learning:
  • Logistic Regression
  • Random Forest Classification
  • Support Vector Classification

Data Source:

1. To set up the data for this project, download the AHS 2017 National survey data here
2. Move the 'household.csv' into the data folder within this repo
3. From the home directory of this repo, import get_data.py and run the create_dataset() function to limit survey data to features related to the neighborhood, as well as the target feature (ADEQUACY)
4. Use the export_dataset() function from the home directory to export cleaned dataset to data folder

Summary and Links to Files:

• Slide Deck (pdf)
• Data Source
• Notebooks
  • Technical Notebook
• Python files