GapSight Health

Mental Health Demand–Supply Gap Monitor (U.S.)

GapSight Health quantifies mental health demand signals (Google Trends) and contrasts them with provider capacity (HRSA/AHRF) to surface demand–supply gaps across U.S. states over time.

Demo: https://mental-health-dashboard-iw4i.onrender.com/

Highlights

• 2020–2024 state–month dataset integrating Google Trends (400+ mental-health queries) + HRSA/AHRF provider-capacity metrics
• Interpretable modeling: XGBoost + SHAP to estimate anxiety/depression demand indices and surface top drivers
• Interactive dashboard to explore gaps and drivers by state and month

Data

Sources

• Google Trends via pytrends (mental-health related queries)
• HRSA/AHRF provider capacity metrics

Note

• Raw extracts can be large and may have redistribution constraints.
• This repo will provide scripts to rebuild the final state–month dataset (WIP).

Repo structure

• app/ dashboard app
• src/ data + modeling pipeline (in progress)
• notebooks/ experiments / analysis
• docs/ report and slides

Attribution / Credits

Personal continuation/refactor of a GWU Capstone team project (Team of 4). Original team members: Erica Zhao, Qibin Huang, Jianjun Gao, Sandhya Karki. Original team repo: QibinHuang/Mental-Health-dashboard

## Modeling (XGBoost)

Training scaffold lives in `src/models/`.

Example (WIP):

```bash
python src/models/train_xgb.py --data path/to/state_month.csv --target anxiety_index --split_date 2024-01

How to run (WIP)

pip install -r requirements.txt
python src/models/train_xgb.py --help

Quickstart (WIP)

pip install -r requirements.txt
python src/models/train_xgb.py --help
python app/app.py