
EP20: Prof Rob Hyndman: Forecasting COVID, time-series, and why causality doesnt matter as much as you think.
The Curious Quant
00:00
Is There a Use Case for Forecasting?
The rate of which you might lose accuracy might mean that again a certain stages of the virus you might be confident in talking about two three weeks ahead but not more. If you provide always provide prediction intervals you soon know what horizon at the you know what point your forecast became useless um because the prediction it was get so wide that they're not telling you anything. One of the mistakes i made in one of my very first forecasting consulting projects was agreeing to forecast further ahead than i should have.
Transcript
Play full episode