The PandemicEstX04.xlsx Model

Estimating a Pandemic Model with What'sBest!

View the model

Download the model

This model estimates a simple "SIR" model that has 4 population segments:

**S**usceptible: Not yet infected, but may be eventually.

** I**nfected: Have the infection.

** R**ecovered: Recovered from infection, not eligible to be re-infected.

Dead: No longer part of total population.

The possible transfers each period are:

S to I: Prob{ Susceptible person gets infected} = alpha * I/(S + I + R) + beta,

where beta is the probability of a spontaneous infection,

so total expected number transferred = alpha * S * I/( S + I + R) + beta * S.

If beta= 0 then Prob{a given susceptible gets infected} is proportional to total population that is infected.

S to D: total transferred is = gamma* S,

I to R: total transferred = gamma * I, so mean time infected = 1/gamma,

I to D: total transferred = delta * S,

R to D: total transferred = zeta * R.

The estimation problem is to find the alpha, beta, gamma, delta, and zeta that best fit the data.
You need What'sBest! from www.lindo.com to solve the estimation/fitting problem.

Keywords:

Bass Model | Diffusion | Multi-period | Estimation | COVID-19 | Contagion | Epidemic | Pandemic | Chain reaction | SIR model |