Video: Katherine J. Igoe
In the vast spectrum of data available to health researchers, policy analysts, economists and clinicians, it can be hard to separate the signal from the noise.
Sherri Rose, HMS associate professor of health care policy, is developing new machine learning tools that she says are bringing “statistical advances for big data and data science to answer critical questions in health economics.” She explains how in this short video.
Instead of telling the machines what to look for and how to find it, machine learning starts by giving the program a goal, then letting the software explore different approaches to analyzing the available data.
These techniques—part of what Rose calls computational health economics—are beginning to show remarkable results: finding better ways to design health insurance plan payment, shifting physician payment incentives to reduce health disparities and even demonstrating that health insurers can use prescription drug utilization data to identify unprofitable enrollees.
