matter

Matter builds 23 machine learning models predicting outcomes across four domains: affordability and family, benefits enrollment, employment, and health.

Each model page shows how to use its score, how it compares with published figures, and how well it works.

To learn what the scores and segments mean and how to build a list from them, start with Finding the people you’re looking for.

A machine-readable catalog of these models, with the same metrics, is at models.json.

0.77mean AUC21 classifiers
0.16mean brier skill21 classifiers
0.02mean decile error21 classifiers

released 2026-10-19 · models as of 2026-10-19 18:01 UTC

models