2
$\begingroup$

I'm working with a SIR infection model, which is $$\begin{array}{rcl} \frac{dS}{dt} & = & -\beta IS\\ \frac{dI}{dt} & = & \beta IS-\gamma I\\ \frac{dR}{dt} & = & \gamma I \end{array}$$ in continuous time, where $S$, $I$, and $R$ are the proportion of Susceptible, Infected, and Recovered, respectively.

However, since I am working with fixed-width discrete-time data, I think it would be more appropriate to modify the equations accordingly. I know this is incorrect (based on getting a negative values for $\hat\gamma$ and $\hat\beta$, neither of which should be negative): $$\begin{array}{rcl} S_{t+1} & = & -\beta I_{t}S_{t}\\ I_{t+1} & = & \beta I_{t}S_{t}-\gamma I_{t}\\ R_{t+1} & = & \gamma I_{t} \end{array}$$

There is a model on Page 2 of this document which incorporates the population $N$, but i'm unsure of the reasoning behind this and doesn't appear to fix my problems in solving for the parameters.

Ultimately, I would like to estimate $\beta$ and $\gamma$ by doing regression on $$I_{t+1} = \beta (I_{t}S_{t})+\gamma (-I_{t})+U_{t} \text{, where }U_{t}\LARGE\text{~}\normalsize N(0,\sigma^2)$$ or whatever the discrete version is.


Edit: I feel as though I should note that I essentially used the discrete version above when solving for the parameters "manually" by looping through several values to get the minimized Mean Square Error. So, I would have expected to be able to just do the regression on the same equation for $I_{t+1}$. Apparently this is not the case as I am getting very different results with the regression.

$\endgroup$
0
2
$\begingroup$

You've missed some things. Remember a derivative is the small $\Delta$ limit of, for example, $(S(t+\Delta)-S(t))/\Delta$. So, your first differential equation is like $$ \frac{S(t+\Delta)-S(t)}{\Delta}=-\beta IS $$ or, where $\Delta$ is your time interval width, $$ S_{t+1}=S_t(1-\Delta\beta I_t). $$ You missed the $S(t)$ term on the right hand side. $S$ should be decreasing with time, and looking at the equation we see that this would be the case if $\beta$ is positive.

Also note that $S+I+R=N$ is constant, so you only need 2 differential equations, you can always find $R$ by doing $n-S-I$.

$\endgroup$
1
  • $\begingroup$ Thank you! This pointed me in the right direction and I got what I needed. $\endgroup$ Oct 19 '15 at 14:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.