Nice article posted by my epidemiology colleague John H Holmes who has been trapped in Northern Italy during this crisis.
Is a nice tutorial explaining the different types of covid models out there, what they are good for and their limitations and uncertainties.
As famous 20th century statistician George Box has said “All models are wrong, some are useful”
That is coming from a lifelong modeler (and posted by one). The key is understanding strengths and limitations and uncertainties of each and using them to inform — they are not crystal balls and should not be followed as such, nor dismissed when they don’t meet that unrealistic high expectation.
Comments