Food Desert Effect on County-Level Health Outcomes

DATA-245 Machine Learning Group Project

Overview

This project investigates how food access inequality correlates with community-level health outcomes (obesity, diabetes) across 2,275 US counties.

Project Structure

ML group project/
├── data/
│   ├── raw/                    # Original source data
│   ├── processed/              # Cleaned datasets
│   └── output/                 # Analysis results (CSV, PNG)
├── notebooks/                  # Jupyter notebooks for analysis
├── src/
│   ├── analysis/               # Core analysis scripts
│   └── utils/                  # Utility modules
├── dashboards/                 # Streamlit applications
├── docs/
│   ├── reports/                # PDF documentation
│   └── images/                 # Charts and diagrams
├── requirements.txt            # Python dependencies
└── README.md

Installation

# Create virtual environment
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

Usage

Run Analysis Scripts

# Regression analysis
python src/analysis/Regression_analysis.py

# PCA analysis
python src/analysis/pcaanalysis.py

Launch Main Dashboard

cd dashboards
streamlit run main_dashboard.py

The dashboard includes 8 pages:

1. Project Overview
2. Data Exploration
3. Regression Analysis (Linear, Ridge, Lasso)
4. Classification Models (Logistic, KNN, Naive Bayes, SVM, Decision Tree, Random Forest, Extra Trees)
5. Clustering Analysis (K-Means, Hierarchical)
6. PCA Analysis
7. Model Comparison
8. Key Insights & Conclusions

Jupyter Notebooks

jupyter notebook notebooks/

Key notebooks:

• Food_Desert_Data_Cleaning.ipynb - Data preprocessing pipeline
• Google_Colab_K_Means.ipynb - Clustering analysis
• Regression_Modeling_Diabetes_Obesity.ipynb - Regression modeling

ML Methods

• Regression: OLS, Ridge, Lasso
• Clustering: K-Means (5 clusters)
• Dimensionality Reduction: PCA

Data

• Source: 2025 County Health Rankings
• Size: 2,275 counties across 48 states
• Target Variables: Adult obesity rate, diabetes rate
• Features: Food environment, socioeconomic factors, education, rurality

Team

DATA-245 Group 3:

• Savitha Vijayarangan - Project Coordination
• Jane Heng - Regression Lead
• Rishi Visweswar Boppana - PCA Lead
• Kapil Reddy Sanikommu - Dashboard Lead