Talk:Compartmental models in epidemiology

Latest comment: 2 years ago by Jaredroach in topic Limitation of models


Limitation of models

edit

Herrmann et al. argue that "none of the above models, currently predicting the COVID-19 pandemic, take into consideration the structure underlying the human interaction network, thus limiting the possible mitigation strategies that can be considered." and "Current COVID-19 models are based on differential equations or random diffusions which assume that human interaction behaviours are generally homogeneous (i.e. alike). It is, however, well-documented that many biological interactions, including human face-to-face communications, are not homogeneous and not random", see https://iopscience.iop.org/article/10.1088/1478-3975/aba8ec - anyone wants to comment or integrate? (Text copied and is available under a Creative Commons Attribution 4.0 International License) --17387349L8764 (talk) 20:34, 3 April 2021 (UTC)Reply

Yup. There are issues with the way this article cites COVID and uses COVID examples. We should strive to make this article a bit more general, and use a variety of examples from many diseases, and limit the COVID examples. Particularly concerning is the Ohio-data figure that shows multiple COVID waves and uses it as an example for the SIR and SIRV model. Neither of these models can predict more than a single wave. So this multi-wave example does not belong as a figure illustrating these models in an encyclopedia article. Jaredroach (talk) 20:03, 20 August 2022 (UTC)Reply