--- tags: Debug --- # Heuristic Debug Rev 5 ### Description #### Heuristic analysis <!-- Optimal mobility restriction = P(infectious) --> P(infectious | evidence) = P(evidence | infectious) P(infectious) / P(evidence) **Behaviour of bdt** = individual +test results in quarantine of individual with +test + quarantine of contacts w/in last 14 days + quarantine households **Behaviour of heuristic** = behaviour of bdt + individual symptoms signal + risk messages caused by test + risk messages caused by symptoms + risk messages about risk For **individual symptoms signal** to be effective, we want to have high p(covid | symptom), and many infections happening between symptom onset and test+ time. **risk messages about symptoms** P(infection | symptom messages) = P(infection) prod_clue P(symptom message | infection) / normalization over infection vs not-infection - \# people exposed to people with symptoms - how informative are symptoms - mismatch time / generation time **risk messages about risk** TODO: - individual symptoms signal - check p(covid | symptom) - check % infections happening between symptom onset and test+ time - risk messages about symptoms - ### Pareto Adoption ![](https://i.imgur.com/wf8V7zR.png) ### Jelly Beans ![](https://i.imgur.com/UrtFDsR.png) ![](https://i.imgur.com/eLFLshM.png) ![](https://i.imgur.com/qWXIzin.png) ### Presymptomatic ![](https://i.imgur.com/YK95b47.png) ![](https://i.imgur.com/9n1e8l8.png) ![](https://i.imgur.com/RflABXa.png) ![](https://i.imgur.com/OPfLvQF.png)