SEANN: A domain-informed neural network for epidemiological insights
Jean-Baptiste Guimbaud, Marc Plantevit, Léa Maître, Rémy Cazabet, Vijay Kumar, Vijay Kumar, Vijay Kumar · 2025
Read SEANN: A domain-informed neural network for epidemiological insights by Jean-Baptiste Guimbaud, Marc Plantevit, Léa Maître, Rémy Cazabet, Vijay Kumar, Vijay Kumar, Vijay Kumar (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-12707678.txt.
Description
Jean-Baptiste Guimbaud and three co-authors present a 2025 neural network, SEANN, built for health studies that respects what the field already knows. Classic tools for linking traits to health are rigid, while plain neural networks are flexible but data-hungry, and health data is often thin. SEANN feeds the network a form of settled knowledge: pooled effect sizes from past meta-studies.
Those pooled sizes enter the training goal itself, steering the network toward links the field already trusts. In controlled simulations, the informed network beat a plain one at forecasting new cases. Its learned links also looked more sound to an epidemiologist's eye. The authors offer the design as a way to make machine learning usable where data is scarce but past studies are rich.
Catalog record
| CREATOR | Jean-Baptiste Guimbaud, Marc Plantevit, Léa Maître, Rémy Cazabet, Vijay Kumar, Vijay Kumar, Vijay Kumar |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 55 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |