Large deviations in non-Markovian stochastic epidemics
Abstract
We develop a framework for non-Markovian SIR and SIS models beyond mean field, utilizing the continuous-time random walk formalism. Using a gamma distribution for the infection and recovery inter-event times as a test case, we derive asymptotical late-time master equations with effective memory kernels and obtain analytical predictions for the final outbreak size distribution in the SIR model, and quasistationary distribution and disease lifetime in the SIS model. We show that increasing memory can greatly widen the outbreak size distribution and reduce the disease lifetime. We also show that rescaled Markovian models fail to capture fluctuations in the non-Markovian case. Overall, our findings, confirmed against numerical simulations, demonstrate that memory strongly shapes epidemic dynamics and paves the way for extending such analyses to structured populations.