Dynamical Survival Analysis: Survival Models for Epidemic
summary
Introduces dynamical survival analysis that replaces the random infectious pressure in a stochastic SIR with its law-of-large-numbers limit, using the Selke construction to get individual infection and recovery time laws, extendable to frailty models for heterogeneity. Proves concentration bounds on the approximation (finite horizon, large N) and shows it handles interval-censored survival data, powers small longitudinal “human sensors” and wastewater surveillance; applied to Ohio/Kentucky cases where model vs serology/wastewater shows roughly a 40% difference from reported cases.
AI-written summary · about