結局、これが核心部分

Rt =I t (number of new infections generated at time t) / Σ[s=1,t] I t-s * Ws ( = total infectiousness of infected individuals at time t)

Ws : an infectivity profile given by a probability distribution ws, dependent on time since infection of the case, s, but independent of calendar time, t.


E[I t] = Rt Σ[s=1,t] I t-s * Ws

Σ[s=1,t] I t-s * Wsの部分が畳み込み積分で

Ws ∝　serial interval　　ガンマ分布で近似するのが定石らしい。