|
We have millions of observations, complex relationships to model, multiple projects to juggle, but limited resources. What can we do? Improvise!
In this case study, we describe using bootstrapping as a expedient and practical solution for modeling the relationship between dependent and independent health care variables which, while linked at the hospital level, are from completely different data sources.
Our discussion isn't limited to methodology. Instead, it's to share the range of responsibilities, challenges, and rewards, of a "data science" team working with health care "big data" and technology.
The hope is that this window into our world will foster deeper and broader involvement of formally trained statisticians with data science teams across industry.
|
