This is the repository for all datasets and Jupyter notebooks utilized in our project. The goal of the project is to provide an allocation prioritization for each county in Pennslyvania for four areas of interest related to improving maternal, infant, and child health (MICH) outcomes. The repository contains the datasets and notebooks used at every stage of our process. The key files to run that provided the main results of our project are in the main folder and explained in the 'Main Files' section below. Data preprocessing steps, raw data sources, and other exploratory notebooks are also stored within other folders in our repository.

To check the results of our project, the main notebook to examine is the 'OptimizationCode.ipynb'. How all other notebooks and files are connected are explained below.

To reproduce the optimization results, run the 'OptimizationCode.ipynb' file. This contains the main and final results of the project.

To reproduce project starting with cleaned data, first run 'LinearRegression.ipynb' followed by 'OptimizationCode.ipynb'. Keep in mind that the results of the regression inform the optimization model. There is a copy of the results already saved in the repository, so the file saving line in the regression notebook is commented out.

To reproduce the project starting with raw data, you must run the preprocessing notebooks before running any of the modeling notebooks. For preprocessing, run 'WebScrapingCode.ipynb', 'Census-ADI.ipynb', and 'DatasetCreation.ipynb' in that order. You will then have the cleaned datasets. Versions of the cleaned datasets are contained in our data folder. After preprocessing is completed, you can run the 'LinearRegression.ipynb' and 'OptimizationCode.ipynb'.

OptimizationCode.ipynb

This notebook contains the optimization implementation of our project. The purpose of the notebook is to determine the prioritization allocation of the programs that were determined from our linear regression model. It references the 'parameters.csv' and 'regression_results.csv'. The final results of our project are contained in this notebook.

LinearRegression.ipynb

This notebook contains our linear regression modeling of our project. The purpose of the notebook is to find what are the most important variables that contribute to achieving MICH goals across the state of Pennslyvania and to obtain coefficient weights to use in our optimization formulation. The notebook contains the entire modeling training and evaluation process. The 'parameters.csv' and 'regression_results.csv' come from this notebook. Please note that these two lines have been commented out because a seed was not intially set with the results that the final optmization results used.

parameters.csv

This csv file contains the county values that are used in our optimization problem. It comes from 'LinearRegression.ipynb'.

regression_results.csv

This csv file contains the coefficient weights that were found from our linear regression model in 'LinearRegression.ipynb'.

This folder contains all the notebooks used to clean and preprocess the data used in our future modeling. It also contains some EDA.

This folder contains the raw and cleaned datasets, in csv form, that were utilized in the project. These datasets are used in the preprocessing and linear regression stages of our project.

This folder contains some other model types, such as CART and classifiers, that were tested in early iterations of the project. These files have no impact on the actual findings of our project.