Koopman Representations for Early Outbreak Warning and Minimal Counterfactual Intervention in Multi-Agent Epidemic Simulations
Florin Leon · 2026
Read Koopman Representations for Early Outbreak Warning and Minimal Counterfactual Intervention in Multi-Agent Epidemic Simulations by Florin Leon (2026) 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/arxiv-2605-01803.txt.
Description
Florin Leon reads the first days of a simulated epidemic for signs of the outbreak to come. He then asks what smallest push could still prevent it. His agents move, differ in susceptibility, and pass infection by co-location, with viral load shaped by immunity, a test bed tuned near the tipping point where small changes flip outcomes.
Daily aggregate readings from early windows are compressed into a Koopman latent space, coordinates in which the messy dynamics run nearly linearly, supporting short forecasts and risk estimates. A random forest on those features predicts whether the final attack rate will cross the major-outbreak line, and the Koopman features drive the class separation. Counterfactual runs search for minimal saves: holding one chosen agent home for one day often cuts the attack rate and pulls the trajectory under the outbreak threshold. The framework joins early warning to intervention choice.
Catalog record
| CREATOR | Florin Leon |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 99 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.01803