Research

Modeling health, one county at a time

Since 2024 I've worked with Dr. Taylor Anderson on computational public health, building models that estimate health outcomes where data is thin, so local decision-makers aren't guessing. The code and datasets are open source.

View my Google Scholar
5.5K synthetic agents

Agent-based models of health behavior

Simulating thousands of synthetic individuals, each with their own social ties and circumstances, to understand how policy and social factors shape decisions like getting vaccinated.

+2% over CDC benchmarks

Small-area estimation

National health surveys are great for states, but blurry at the county or neighborhood level. I work on methods (like iterative proportional fitting and spatial microsimulation) that sharpen those local estimates.

First-author paper

Fast surrogate models

Full simulations are slow. Geographically weighted surrogate models can approximate them quickly, making chronic-disease estimates practical for decision-makers.

Publications

Peer-reviewed papers

  1. Accepted

    Geographically Weighted Surrogate Models for Rapid Small-Area Chronic Disease Estimation

    Gupta, A., Péter, S. A., et al.

    Applied Geography · accepted, in press

  2. Accepted

    A Synthetic Population and Small-Area Estimate Dataset for Health Behaviors and Outcomes in the United States

    Von Hoene, E., Gupta, A., et al.

    Nature Scientific Data · accepted, in press

  3. 2026

    Evaluation of a spatial microsimulation framework for small-area estimation of population health outcomes using the Behavioral Risk Factor Surveillance System

    Von Hoene, E., Gupta, A., Kavak, H., et al.

    International Journal of Health Geographics

    DOI →
  4. 2025

    Evaluating the Transferability of a Synthetic Population Generation Approach for Public Health Applications

    Von Hoene, E., Gupta, A., et al.

    Winter Simulation Conference

    DOI →

Citations and the most up-to-date list live on Google Scholar.