Although this article doesn’t go into much detail about what sort of evaluation metric was used or what the results of the model happen to be, this paper does. This article was written almost 20 years ago and goes into the sort of a detail a Data Scientists or Machine Learning Researcher would appreciate.
Cutting-edge data science can help address many of the serious challenges our healthcare systems are facing today and in the future. Here we look at how big data analytics and machine learning can help pull together various data sources to predict admission rates in hospitals.
It’s interesting to consider as well the length of the prediction, “15-days”. Most weather forecasting system at present do 10-days as an upper bound and only recently moved from 5-days to 7-days back in 2000. Primarily this has to do with the decaying quality of the prediction over that period of time. So the question what is special about “15-days” for hospital admissions comes to mind. All in all, we are excited to see machine learning being used in healthcare.