Main content start

A more honest model

Or, according to Emmanuel Candès, combining quantile regression and conformal prediction has produced "the most informative, well-calibrated range of predicted values that I know how to build."

In an article that profiles the research of Candès, recent postdoctoral scholar Yaniv Romano, former graduate student Evan Patterson as well as current PhD student John Cherian, today's Stanford Report describes this model as the first real-world application of an existing statistical technique developed at Stanford. The technique is applicable to a variety of problems, including to data about COVID-19 survival times, and could help elevate the importance of honest uncertainty in forecasting.