Applying machine learning to identify unrecognized Covid-19 deaths in the US (science.org)

> Our machine learning approach relies on different assumptions—(i) that there was no measurement error for COVID-19 deaths in hospitals and (ii) that a model trained on in-hospital deaths is transportable to out-of-hospital settings.

Whether these assumptions are valid is highly debatable given various incentive structures applied in in-hospital settings over the period studied.

I thought we needed a ML model to deflate Covid deaths. In the US at least, a lot of died "with Covid" were counted as died "of Covid"