Virginia Population Modeling by Elliot Hong & Thomas Welch

This project uses machine learning to predict annual population estimates across Virginia using demographic data.

Models

• Linear Regression (baseline)
• Random Forest Regressor
• LSTM Neural Network

Methods

• One-hot encoding, scaling, feature selection
• 5-fold CV with RMSE and R² evaluation

Key Results

• Random Forest had best performance (RMSE ≈ 266)
• LSTM struggled due to limited temporal data
• Top features: age, year, locality, race

Data

Virginia Single-Race Population Estimates

Files

• VA_Population_Modeling.ipynb: modeling notebook
• template.tex: final report (LaTeX)
• figures/: plots