An Assessment of Poverty and Possible Correlates

Objective:

• The ambition of this project was to build a basic analysis of poverty with a special emphasis on poverty within the United States.

The Work:

• Presentation is available at (placeholder html).
• CSV data available under csv folder. The csv "city_pov_2016_mn.csv" is a large csv that includes all possible variables to analyze with poverty levels, and was used to find, then select, correlation coefficients for regression and model building. "Main.csv" is the csv that included the variables that we selected to build the models.
• The folder "main_notebook_folder" includes the jupyter notebooks used for putting together our analysis.
• "Model_data_explore" folder includes jupyter notebooks made specifically for data exploration.
• "Pages" is the location of the html forms of our jupyter notebooks found in the "main_notebook_folder."

Data:

• United States Census Bureau (https://factfinder.census.gov)
• World Bank Databank (https://data.worldbank.org)
• Center for Disease Control and Prevention, 500 Cities (CDC, https://chronicdata.cdc.gov)

Tools:

• Leaflet.js
• Mapbox
• jupyter notebook
• pandas, numpy, scipy
• sci-kit learn for:
  • Linear Regressions
  • Lasso, Ridge Regressions
  • Linear SVC
  • K Neighbors Classifier
  • Random Forest
  • Logistic Regression
  • Decision Tree