SL-BiLEM: A New Epidemic Model Integrating Human Behavior Feedback
Researchers have developed SL-BiLEM (Structured Learnable Behavior-in-the-Loop Epidemic Model), a novel framework that addresses the challenge of distribution shifts in epidemic forecasting caused by human behavioral responses to disease spread. The model decomposes effective transmission into components including baseline transmissibility, policy effects, media influence, and compliance, with physical constraints like monotonicity and smoothness ensuring robust extrapolation under new policy regimes. SL-BiLEM also enables counterfactual analysis for intervention decision support. Validation was conducted on three real-world datasets: a cruise ship outbreak, school influenza, and school-dis. The work is published on arXiv (2605.26704).
Key facts
- SL-BiLEM stands for Structured Learnable Behavior-in-the-Loop Epidemic Model.
- It addresses distribution shifts at policy intervention points due to human behavior feedback.
- The model decomposes effective transmission as β_eff(t,g) = β_0(g) × m_policy(t) × m_media(t) × m_comp(t,g).
- Physical constraints include monotonicity, smoothness, and bounded-jump on the compliance function.
- SL-BiLEM supports counterfactual analysis for policy evaluation.
- Validation used three real-world datasets: cruise ship, school influenza, and school-dis.
- The paper is available on arXiv with ID 2605.26704.
- The approach aims to improve predictive validity under novel policy regimes.
Entities
Institutions
- arXiv